Two helices from one chiral centre – self organization of disc shaped chiral nanoparticles

Gold nanoparticles (AuNPs) have been prepared and surface-functionalized with a mixture of 1-hexanethiol co-ligands and chiral discogen ligands separated from a disulfide function via a flexible spacer. Polarized optical microscopy together with differential scanning calorimetry showed that the organic corona of the nanocomposite forms a stable chiral discotic nematic phase with a wide thermal range. Synchrotron X-ray diffraction showed that gold NPs form a superlattice with p2 plane symmetry. Analysis indicated that the organic corona takes up the shape of a flexible macrodisk. Synchrotron radiation-based circular dichroism signals of thin films are significantly enhanced on the isotropic-LC transition, in line with the formation of a chiral nematic phase of the organic corona. At lower temperatures the appearance of CD signals at longer wavelengths is associated with the chiral organisation of the NPs and is indicative of the formation of a second helical structure. The decreased volume required and the chiral environment of the disc ligands drives the nanoparticles into columns that arrange helically, parallel to the shortest axis of the two dimensional lattice.

Gold nanoparticles (NPs) were surface functionalized with hexanethiol groups and chiral nematic discogen ligands. A superlattice, liquid crystal behaviour and helix formation of the discs and a second helical organisation of the NPs were detected.

The exploration of chiral nanomaterials is a rapidly evolving fascinating field with uses ranging from catalysis, chiral molecular sensing, stereoselective separations to super-resolution imaging.^1–3 Bottom-up approach design strategies for nanoparticles (NPs) or nanorods (NRs) as composites in liquid crystal (LC) matrices have been developed, making use of progress in synthesis strategies and characterization techniques.^4–6 For optical applications symmetry breaking by introducing chirality into the systems at the nanoscale is critical.⁷ So far, efforts to obtain chiroptical NP-LC systems have relied mainly on NPs doped in LC hosts: either chiral nematic templates disperse NPs into chiral assemblies⁸ or the NPs functionalized with chiral ligands are included in achiral nematic LCs;⁹ both strategies aim to transfer and amplify chiral effects.^10,11 Beyond the scientific challenge of generating nanocomposites where the interplay of bottom-up structuring enables plasmonic interactions of the NPs, leading to novel, so called metamaterials properties, there are new uses ranging from novel display devices,¹² to more advanced optical communications using spin information between LEDs¹³ and in quantum teleportation for optical computing.¹⁴Precise size control of NPs in LC matrices in combined systems has been found critical to avoid phase separation or precipitation of NPs.¹⁵ For intrinsic liquid crystalline NP-LC systems, targeting of the LC phase range to be close to ambient is of utmost importance, as delamination of ligands occurs at elevated temperatures, even for very stable thiol functions. Degradation tends to occur above 100 °C.¹⁶ Inclusion of chiral groups in the organic corona can lead to chiral LC phase behavior, but there is little evidence that the NPs experience a chiral distortion in their packing; experimental evidence indicates that the nanocomposites minimize energy by packing in non-chiral superlattices and chirality is confined to the corona.¹⁷Here we show that a combination of small NPs with a carefully designed chiral organic coating results in chiral assembly behaviour of both the NPs and the organic corona. Our approach is strictly modular. The surface of the gold NPs is covered and passivated by hexylthiol groups and chiral lipoic acid derivates bearing via a spacer the mesogenic groups. We note that lipoic acid, a naturally occurring chiral antioxidant explored elsewhere for the treatment of several diseases, is bound with a dithiolate bridge via two atoms to the Au surface (see Chart 1).¹⁸ This places the chiral moiety rigidly in position onto the NPs, making thus a chiral surface.¹⁹ A pentaalkynylbenzene (PA) derivative was selected as mesogenic group, as this structural motif is known to promote reliably discotic nematic phase (N_D) behaviour.²⁰ The much larger volume of the PA group, when compared to typical rod shaped mesogens, whilst also delivering LC behaviour close to ambient, is anticipated to impact more strongly on the spatial organization of neighbouring NPs when compared to calamitic functionalized NP systems.¹⁷ The formation of helically ordered superstructures is studied by a combination of dedicated thin film optical polarizing microscopy, X-ray diffraction and synchrotron radiation-based circular dichroism (SRCD). A structural model is proposed based on a chiral nematic phase made up of AuNPs coated with chiral shells forming flexible macrodisks which in turn assemble in a superlattice.^21–23Open in a separate windowChart 1Sketch of the functionalized gold nanoparticle system AuDLC*.The design, synthesis and chemical characterization of the chiral discogen is described in the ESI.^† The precursor disk-like mesogen which forms exclusively a N_D phase above 100 °C consists of six aromatic rings flanked by flexible pentyloxy chains and an undecyloxy chain bearing a terminal hydroxy group (Fig. S1^†). The attachment of the flexible (R)-(+)-1,2-dithiolane-3-pentanoic tail was predicted to promote the desired phase. The new target discogen was obtained in a high yield (93.1%) reaction. For the synthesis and purification of LC-functionalized AuNPs, denoted as AuDLC*, a synthetic pathway was developed involving first the preparing and purifying of the alkylthiol coated NPs to which LC groups are linked, using an exchange reaction (for details see the Scheme 1 in ESI^†).^24,25The thermal behaviour of the chiral discogen was examined by differential scanning calorimetry (DSC) and polarized optical microscopy (POM). The target discogen shows clearly a chiral nematic phase with the characteristic bundles of oily streaks (Fig. S2^†). On heating (Fig. S3^†), phase formation was observed at 49.0 °C from the crystalline state, with the material turning into an isotropic liquid at 95.7 °C (ΔH = 0.20 J g⁻¹ or 0.28 kJ mol⁻¹). Formation of the mesophase from the isotropic state on cooling started from 94.5 °C with an enthalpy change of 0.12 J g⁻¹ (0.17 kJ mol⁻¹), followed by a slow crystallisation occurring at 5.1 °C. In brief, the chiral phase behaviour of the target ligand is exclusively chiral nematic, enantiotropic and with a wide thermal range of up to 89.4 °C.The ¹H-NMR spectra of the investigated AuDLC* nanocomposite and the comparison with the spectra of the monomer ligand (Fig. S4^†) indicate that the chiral discogen with the disulfide group are chemically attached onto the surface of AuNPs with high efficiency. A typical feature for such NPs are the broadened ¹H-NMR spectra caused by the reduced mobility of alkyl chains of the ligands when anchored to NPs and additionally the sharp peaks present in the free ligand have disappeared, typical for such systems.²⁶ The absence of mobile free discogen using thin layer chromatography (silica, with dichloromethane as mobile phase) confirms this result. The amount of chiral discogen bounded onto the gold surface relative to the number of 1-hexanethiol groups was determined from the ¹H-NMR spectra of AuDLC*. The ratio of hydrocarbons to discogens was found to be 2 : 3. Transmission Electron Microscopy (TEM) investigations confirm that this sequential synthesis affords nanoparticles of low polydispersity (Fig. 1a); the average particle size is ∼2.5 ± 0.5 nm. Thermo-gravimetric analysis (TGA) was performed (Fig. S6^†) and gives the weight fraction of gold in AuDLC* to be 47.8% (wt/wt). The mesogen density per Au NP was calculated (ESI^† Part 7) by combining the ¹H-NMR results (Fig. S4^†) with the model developed by Gelbart et al.²⁷ On the basis of these data, we obtain that there are on average 211 Au atoms per particle, and a total number of 53 ligands per particle, of which 32 are chiral discogens and 21 are hexylthiol groups (Table S1^†). This is in accord with a ligand density of up to ∼6.3 ligands per nm².²⁸ Additionally, the characteristic peaks in the UV-vis spectra of AuDLC* arise from the combination of gold (∼500 nm) and LC ligands (absorption maxima at 238, 261 and 351 nm with two shoulder peaks at 378 and 421 nm) (Fig. S7^†).Open in a separate windowFig. 1(a) TEM image of 2.5 nm AuDLC* nanocomposite (inset: the size distribution of AuDLC*). POM micrographs of AuDLC* (b) 86.5 °C (90° crossed polarizer) (×100 μm). (c) 37.8 °C (90° crossed polarizer) (×100 μm). (d) DSC of AuDLC* at heating and cooling rate of 10.0 °C min⁻¹ (e) SAXS diffractograms of AuDLC* on heating from 30 °C to 140 °C. (f) SRCD spectra of sheared AuDLC*. Recorded in 5 °C steps on cooling from 100 °C to 30 °C. The enlarged plots in dotted regions A–C are shown in Fig. S12b–d.^† The CD intensities at a selected wavelengths in each region (A, B and C) are plotted as a function of temperature in Fig. S13^† and these plots clearly show the difference in the formation process between the two helical structures.The mesophase behaviour of the AuDLC* nanocomposite was first characterized by POM observation. As shown in Fig. 1b and c, upon cooling from the isotropic liquid, the colour in the POM texture changes from light green to yellow and we associate this with the helical pitch in AuDLC* increasing with decreasing temperature. In some regions of the POM slides, a chiral nematic phase with the typical Grandjean texture (Fig. S8a and b^†) is found. The LC textures in AuDLC* do recover after gently pressing the microscope coverslip indicating that the LC-decorated NPs form a stable phase and birefringent textures remain stable at ambient (Fig. S9^†). DSC investigations (Fig. 1d) confirm that the AuDLC* nanocomposite shows a thermodynamically stable (enantiotropic) mesophase. In DSC experiments on cooling a small peak associated with the isotropic to phase transition (92.2 °C) can be detected, and at a low temperature (3.8 °C), the glass transition occurs. The transition peak for the transition is wider than in the pure chiral discogen, typical for such systems.⁴ This is associated with the reduced mobility of the LC groups attached to the NPs. The thermal transitions determined by DSC for the pure chiral discogen and AuDLC* are collected in Table 1, showing the low values typical for transitions.²⁰ In contrast to the free chiral ligand, the NPs coated with the chiral ligand and hexylthiol groups show a small decrease in transition temperatures. This could be attributed to both the plastifying effects of the hexylthiol co-ligands and packing constraints of the LC groups due to the attachment to the NPs. This view is supported by the DSC data for the transition at 0.043 J g⁻¹ (0.12 kJ mol⁻¹, considering the evaluated 32 chiral discogen ligands per particle).Transition temperatures (°C) of free chiral discogen and AuDLC*, as determined by DSC (second cooling at rate of 10.0 °C min⁻¹)^a

Electrical conductivity in a non-covalent two-dimensional porous organic material with high crystallinity

Electroactive macrocycle building blocks are a promising route to new types of functional two-dimensional porous organic frameworks. Our strategy uses conjugated macrocycles that organize into two dimensional porous sheets via non-covalent van der Waals interactions, to make ultrathin films that are just one molecule thick. In bulk, these two-dimensional (2D) sheets stack into a three-dimensional van der Waals crystal, where relatively weak alkyl–alkyl interactions constitute the interface between these sheets. With the liquid-phase exfoliation, we are able to obtain films as thin as two molecular layers. Further using a combination of liquid-phase and mechanical exfoliation, we are able to create non-covalent sheets over a large area (>100 μm²). The ultrathin porous films maintain the single crystal packing from the macrocyclic structure and are electrically conductive. We demonstrate that this new type of 2D non-covalent porous organic framework can be used as the active layer in a field effect transistor device with graphene source and drain contacts along with hexagonal boron nitride as the gate dielectric interface.

Ultrathin porous films held together by non-covalent van der Waals interactions was obtained by a top-down approach, which is then utilized as channel material in a two-dimensional planar field-effect transistor device through easy stamp transfer.

We describe a new type of two-dimensional (2D), molecularly-thin porous organic framework that is formed from macrocyclic building blocks that assemble, through non-covalent interactions, into a porous two-dimensional plane. Covalent organic frameworks (COFs) are promising in applications due to their ability to host other functional molecules in the voids.^1–7 Many porous frameworks have been demonstrated to be useful in energy storage,⁸ catalysis,^9–11 separation,^12,13 optoelectronics^4,14 and sensing.^15,16 In order to construct nanodevices with porous channels, ultrathin films of porous frameworks has been prepared with bottom-up^4,17–20 and top-down^1,21 approaches. The top-down approaches to these materials are enabled by strong covalent bonds in the two-dimensional plane and weak van der Waals interactions between them, similar to what is seen in two-dimensional materials such as graphene and TMDs.^22–27 For porous ultrathin films, the electrical conductance has not been extensively investigated.^2,7,20,28,29Here, we explore making molecularly thin layers in which conjugated macrocycles are used as building blocks and non-covalent van der Waals interactions are the adhesive that assembles these molecules into rigid, porous layers. By adjusting the relative strengths of the interactions that direct the assembly within the plane and those holding the two-dimensional layers with respect to each other, we can exfoliate these non-covalent porous frameworks using the same means employed for traditional two-dimensional van der Waals materials.³⁰ Using liquid-phase and mechanical exfoliation, we create porous films that are as thin as two-layers of molecules. These new results are exciting and useful because previously we were not able to obtain such high-ordered thin porous film directly from its bulk crystal and were limited to investigating the electronic properties of this hollow organic capsules in spin-coated films. These ultrathin porous films are ordered over large areas and maintain the single crystal packing from the macrocyclic building blocks. To demonstrate the utility of this new type of ultrathin material, we fabricated 2D field effect transistor (FET) devices in which graphene is the source/drain contacts, hexagonal boron nitride is the gate dielectric interface, and the exfoliated molecular sheet is the active layer. These ultrathin self-assembled materials are efficacious at transporting electrons and will find utility in gas sensing and applications similar to traditional two-dimensional materials. Fig. 1 displays the molecular building block (1). Characterization is contained in the ESI^† and a previous report.³¹1 has several important molecular features in its solid-state assembly. It is a rigid and shape persistent macrocycle that has an interior and an exterior (Fig. 1a), and in bulk, has a pore of ∼11.4 Å in diameter and a surface area of 20 m² g⁻¹ from BET measurements.³¹ When it assembles in the crystalline state, it forms two-dimensional porous sheets with two types of cavities (Fig. 1b), one molecule thick, that are held together by relatively strong π–π contacts and Br–PDI interaction between the bromine atoms on the thiophenes and adjacent PDI molecules (Fig. 1c), which plays a crucial role in the self-assembly of the films. The close proximity of the molecules in the 2D plane together with the conjugation within the macrocycle facilitate charge transport of electrons in the 2D plane. These electrically conductive porous sheets then stack into a three-dimensional crystal in which adjacent sheets are separated from one another by the alkane sidechains of the perylene diimide (Fig. 1d). It is in this alkane gallery that we see an opportunity for exfoliation to yield ultrathin 2D sheets.Open in a separate windowFig. 1Structure and packing of molecule 1. (a) Chemical structure, side view and top view of molecular structure of 1. C, N, O, S, Br are colored in grey, blue, red, yellow and green, respectively. The vertical distance of one macrocycle is about 1.5 nm. (b) Face-on view and edge-on view of one layer of 1. The internal cavity of 1 and the cavity formed by the packing of 1 are labeled as i and i′, respectively. (c) Interactions binding two adjacent macrocycles from neighboring brominated thiophene rings. (d) View of packing of 1 along the c axis through the interaction of alkane sidechains. Exfoliation takes place at the alkyl–alkyl interface. One layer of 1 is about 2 nm in thickness.We isolated crystals of this material that were grown from solution and then tested whether they can be exfoliated. Fig. 2a displays a representative micrograph of one of the crystals. The crystal has a pseudo-hexagonal packing of the molecular building blocks in the two-dimensional plane, and this symmetry is mirrored in the hexagonally-shaped crystals. The simplest method for exfoliation is mechanical exfoliation, which is most commonly performed using scotch-tape.^32,33 We place the single crystals onto clean scotch-tape and repeat the mechanical exfoliation process for a few repetitions, and then we transfer the exfoliated crystals onto a clean silicon wafer. Fig. 2b displays an atomic force microscopy (AFM) micrograph of the typical non-covalent porous 2D sheet of 1 we obtained from this method. The porous sheets are flat and smooth and a few micrometers in diameter with a thickness of ∼8 nm; this thickness corresponds to a stack of five molecular layers of 1. This result demonstrates that non-covalent interactions are strong enough to hold molecules together to form ultrathin porous materials. Just as with traditional two-dimensional materials, the non-covalent porous organic 2D sheets of 1 are flexible as evidenced by the wrinkles and folds in the micrograph in Fig. 2b and S2.^†Open in a separate windowFig. 2Mechanical exfoliation of 1. (a) Optical microscopy and scanning electron microscope (SEM) (inset) images of a single crystal of macrocycle 1. (b) AFM and optical microscope (inset) images of the ultrathin non-covalent porous sheet of 1 on a silicon wafer obtained from mechanical exfoliation.We were unable to obtain porous films as thin as a single layer and also large-area samples using mechanical exfoliation, and thus we next explored if liquid-phase exfoliation^34,35 could produce them. Because the halogen bonds that hold the sheets together should be most robust in solvents with a low-dielectric constant that lack heteroatoms, and because the groups holding the sheets together are the alkane sidechains, we chose saturated hydrocarbons (hexane or heptane) as the solvents for exfoliation. Fig. 3a shows the process we follow for the liquid-phase exfoliation. We suspend single crystals of 1 in heptane and sonicate the mixture for five minutes. We drop cast the supernatant solution on silicon wafer and examine them with AFM. Remarkably, we are able to obtain non-covalent porous organic frameworks as thin as only two molecular layers (Fig. 3b).³⁶ Nevertheless, the lateral size of the porous 2D sheets of 1 we could obtain using this method are quite small, making it difficult to fabricate devices from them.Open in a separate windowFig. 3Liquid-phase exfoliation and combination of liquid-phase and mechanical exfoliation. (a) Schematic showing the liquid-phase exfoliation process. (b) AFM image of the ultrathin sheet of 1 on a silicon wafer obtained from liquid-phase exfoliation method. The sheets in this micrograph are two molecular layers (left) and three molecular layers (right) in thickness. (c) AFM image of the ultrathin sheet of 1 on a silicon wafer obtained from combination of liquid-phase and mechanical exfoliation, inset: optical microscope image of the ultrathin sheet 1. (d) AFM image showing the height change across the ultrathin sheet of 1 on silicon wafer obtained with combination methods, inset: optical microscope image of the ultrathin sheet of 1.To get large-area films characteristic of the mechanical exfoliation and thin films characteristic of the liquid-phase exfoliation, we combined the two methods. We first immerse the crystal of 1 in heptane for a few minutes to let the solvent seep into the gallery between the sheets and weaken the interlayer interactions. Then, we use mechanical exfoliation to isolate the ultrathin films. With this method, we obtained sheets of 1 with a lateral size of over ten micrometers, as shown in Fig. 3c. By carefully examining the exfoliated non-covalent porous 2D sheets of 1, we were also able to observe the height change across the sheet (Fig. 3d) with integer values of the layer thickness after the exfoliation steps. As marked red in Fig. 3d, we could identify a single layer of 1 with a height difference between these two surfaces of about 1.5 nm, corresponding to monolayer of molecule 1.We conducted transmission electron microscopy (TEM) studies to characterize the crystallinity of the as-prepared non-covalent porous 2D sheets of 1. As shown in Fig. 4a (inset), the 2D sheets exhibit a layered structure after liquid-phase exfoliation. The selected area electron diffraction (SAED) in the area (marked by the red circle) reveals a hexagonal diffraction pattern, with the bright reflections corresponding to the (2 1(−) 0) plane, with a spacing of 11.3 Å. This diffraction pattern confirms that the non-covalent, porous 2D sheets of 1 retain the single crystal packing and are stable to the liquid-phase processing.Open in a separate windowFig. 4TEM characterization and device fabrication. (a) SEAD pattern and TEM image (inset) of the non-covalent porous ultrathin sheet 1 obtained by liquid-phase exfoliation. (b) Schematic showing the structure of the hBN/Graphene/1 device based on the non-covalent porous ultrathin sheet 1 with graphene as electrodes and hBN as the dielectric layer. (c) Optical microscope image showing the as-fabricated hBN/Graphene/1 device. (d) Transfer curve of the hBN/Graphene/1 device.We next sought to determine the ability of these non-covalent porous ultrathin layers to transport charge. Because these films are van der Waals materials, we sought to make devices with van der Waals interfaces. [The ESI^† contains the current/voltage curves for 1 in a more traditional organic FET with Au contacts and trichloro(octadecyl)silane coated SiO₂ as the gate dielectric.] The source-drain contacts were fabricated from graphene and the dielectric interface was hexagonal boron nitride (hBN). A schematic of the device is shown in Fig. 4b. To create this device, we first exfoliate graphene and hBN onto a silicon substrate. We then follow a published procedure³⁷ to first pick up hBN and then graphene to make the hBN/graphene stack. We transfer this hBN/graphene stack onto another clean silicon substrate. Then the graphene was cut using electron beam lithography and an oxygen plasma to open a 300 nm gap between graphene electrodes (see Fig. S4^† for the AFM details of graphene electrodes). In order to transfer the non-covalent porous ultrathin sheets of 1 onto the graphene electrodes, we exploit the combined liquid/mechanical exfoliation method above to obtained 2D sheet of 1 with polydimethylsiloxane (PDMS) polymer as the substrate, which was then used for the stamp transferring. In this manner, we were able to transfer the ultrathin sheets (∼20 nm) onto the graphene electrodes. Fig. 4c displays the optical microscope image of the device, and Fig. 4d displays the FET transfer curves revealing that the material is an efficacious, n-type transistor. Several features of the device are noteworthy. The electron mobility in the linear regime was estimated to be 1.6 × 10⁻⁴ cm² V⁻¹ s⁻¹ from the transfer curve. As expected, it is somewhat lower than the electron mobility estimated from the saturation regime of the traditional OFET shown in Fig. S3.^†38 Despite the small size and the nanoscale thickness, the material exhibits over 3 orders of magnitude difference in current between the off and on state of the device. The threshold voltage is about 39 V, implying that the device turns on at relatively high voltage. We surmise that the contact, through the alkane sidechains is an impediment to more efficient charge transport.     

Pressure-and temperature induced phase transitions,piezochromism, NLC behaviour and pressure controlled Jahn–Teller switching in a Cu-based framework

In situ single-crystal diffraction and spectroscopic techniques have been used to study a previously unreported Cu-framework bis[1-(4-pyridyl)butane-1,3-dione]copper(ii) (CuPyr-I). CuPyr-I was found to exhibit high-pressure and low-temperature phase transitions, piezochromism, negative linear compressibility, and a pressure induced Jahn–Teller switch, where the switching pressure was hydrostatic media dependent.

In situ high-pressure single-crystal diffraction and spectroscopic techniques have been used to study a previously unreported Cu-framework bis[1-(4-pyridyl)butane-1,3-dione]copper(ii) (CuPyr-I).

High-pressure crystallographic experiments over the last 25 years or so have proved to be a unique tool in probing the mechanical properties of the organic solid state,¹ metal-complexes, and 2D/3D coordination compounds.² In particular, high-pressure techniques have been used to study an array of mechanical and chemical properties of crystals, such as changes in electrical and thermal conductivity,³ pressure-induced melting,⁴ solubility,⁵ amorphisation,⁶ post-synthetic modification,⁷ and chemical reactions such as polymerisation,⁸ cycloaddition⁹ and nanoparticle formation.¹⁰ Previous high-pressure experiments on porous metal-organic framework (MOF) materials have shown that on loading a diamond anvil cell (DAC) with a single-crystal or polycrystalline powder, the hydrostatic medium that surrounds the sample (to ensure hydrostatic conditions) can be forced inside the pores on increasing pressure, causing the pore and sample volume to increase with applied pressure.¹¹ This technique has also been used to determine the position of CH₄ and CO₂ molecules inside the small pores of a Sc-based MOF at room temperature using a laboratory X-ray diffractometer, and has proved useful in experimentally determining the maximum size of guest molecules that can penetrate into a pore.¹²On direct compression of more dense frameworks, negative linear compressibility (NLC) has also been observed, which results in an expansion of one or more of the unit cell dimensions with an overall contraction in volume. Such changes in the compressibility behaviour of metal-containing framework materials is usually as a result of common structural motifs which rotate or bend in order to accommodate increases in length along particular crystallographic directions.¹³ Changes in coordination environment can also be induced at pressure, as metal–ligand bonds are more susceptible to compression than covalent bonds.^2a In previous high pressure studies on metal complexes or coordination compounds, in which the metal ion has an asymmetric octahedral environment caused by Jahn–Teller (JT) distortions for example (such as those observed in Cu²⁺ and Mn³⁺ complexes), the application of pressure can result in compression of the JT axis, and can even be switched to lie along another bonding direction within the octahedron.¹⁴ Such distortions often result in piezochromism, often observed within a single crystal.¹⁵Here, we present a high-pressure crystallographic study on a novel and unreported Cu-framework bis[1-(4-pyridyl)butane-1,3-dione]copper(ii) (hereafter referred to as CuPyr-I). On application of pressure, CuPyr-I is highly unusual in that it demonstrates several of these phenomena within the same framework, including a single-crystal to single-crystal phase transition, a switching of the JT axis that depends on the hydrostatic medium used to compress the crystal, piezochromism and NLC behaviour. To date, we are unaware of any other material which exhibits all of these phenomena, with the first ever reported hydrostatic media ‘tuneable’ JT-switching.Under ambient temperature and pressure CuPyr-I crystallises in the rhombohedral space group R$\bar{3}$ (a/b = 26.5936(31) Å, c = 7.7475(9) Å). Each Cu-centre is coordinated to four 1-(4-pyridyl)butane-1,3-dione linkers, two of these ligands are bound through the dione O-atoms, with the final two bonding through the N-atom of the pyridine ring to form a 3D polymer. The crystal structure of CuPyr-I is composed of an interpenetration of these 3D polymers to form one-dimensional porous channels (∼2 Å in diameter) that run along the c-axis direction (Fig. 1).Open in a separate windowFig. 1Ball and stick model showing the coordination environment around the Cu²⁺ ion in CuPyr-I, and 3D-pore structure as viewed along the c-axis direction. The yellow sphere represents the available pore-space. Colour scheme is red: oxygen, blue: nitrogen, black: carbon, white: hydrogen and cyan: copper. The Cu²⁺ octahedron is illustrated in green.On increasing pressure from 0.07 GPa to 1.56 GPa using Fluorinert FC-70 (a mixture of large perfluorinated hydrocarbons) as a hydrostatic medium, compression of the framework occurs, resulting in a 9.89% reduction in volume, while the a/b-axes and c-axis are reduced by 4.46% and 1.25% respectively (Fig. 2 (blue triangles) and Table S1^†).Open in a separate windowFig. 2 a/b and c-axes as a function of pressure in a hydrostatic medium of FC-70 (blue triangles) and MeOH (red/black circles). The vertical line indicates the transition from CuPyr-I (red circles) to CuPyr-II (black circles) above 2.15 GPa. Errors in cell-lengths are smaller than the symbols plotted.On increasing pressure to 1.84 GPa, the framework became amorphous, though this is unsurprising as the hydrostatic limit for FC-70 is ∼2 GPa, and compression of frameworks in non-hydrostatic conditions usually results in amorphisation.¹⁶ On increasing pressure to 1.56 GPa, the three-symmetry independent Cu–O/N bond lengths to the ligand were monitored (Fig. 3 and Table S5^†). Under ambient pressure conditions, the two Cu–N1 pyridine bonds are longer than the four Cu–O1/O2 dione bonds, typical for an elongated JT distorted Cu²⁺ complex. However, on increasing pressure the direction of the JT axis gradually changed from Cu–N1 to the Cu–O1 bond (the dione oxygen in the 3-position), becoming equidistant at ∼0.57 GPa. By 1.56 GPa, the lengths of the Cu–N1 and Cu–O1 bonds had steadily reduced and increased by 12.3% and 8.9%, respectively. Throughout this the Cu–O2 bond remained essentially unchanged.Open in a separate windowFig. 3Cu–O1 (orange), Cu–N1 (blue) and Cu–O2 (green) bond lengths on increasing pressure in both FC-70 (triangles and dashed lines) and MeOH (circles and solid lines).Pressure induced JT switching has been observed in other systems, including a Mn₁₂ single-molecule magnet cluster that re-orientates the JT axis on one of the Mn centres at 2.5 GPa.^14a A similar transition was also observed in [CuF₂(H₂O)₂(pyz)] (pyz = pyrazine) and Rb₂CuCl₄(H₂O)₂,¹⁵ where the JT axis was reoriented from the Cu–N bond to the perpendicular Cu–O bond, though this occurs during a crystallographic phase transition at 1.8 GPa.¹⁸ Here, in CuPyr-I, no phase transition takes place, and unusually the JT switching appears to occur progressively on increasing pressure with no phase transition.^14bUsing methanol (MeOH) as the hydrostatic medium, CuPyr-I was compressed in two separate experiments, from 0.52 GPa to 5.28 GPa using synchrotron radiation, and from 0.34 GPa to 2.95 GPa using a laboratory X-ray diffractometer. On increasing pressure to 2.15 GPa, the a/b and c-axes compressed by 6.22% and 0.39% respectively (Fig. 2, Tables S2 and S3^†). On increasing pressure from 2.15 GPa to 2.78 GPa, CuPyr-I underwent a single-crystal to single-crystal isosymmetric phase transition to a previously unobserved phase (hereafter referred to as CuPyr-II).The transition to CuPyr-II resulted in a doubling of the a/b-axes, whilst the c-axis remained essentially unchanged. On increasing the pressure further, the a/b-axes continued to be compressed, whilst the c-axis increased in length, exhibiting negative linear compressibility (NLC) until the sample became amorphous at 5.28 GPa. The diffraction data were of poor quality after the phase transition, and only the connectivity of the CuPyr-II phase could be determined at 3.34 GPa. Above 3.34 GPa, only unit cell dimensions could be extracted. The occurrence of positive linear compressibility (PLC) followed by NLC is unusual in a framework material, and we could find only a few examples in the literature where this occurs.¹⁹During the NLC, the c-axis expanded by 1.46%, to give a compressibility of K_NLC = −5.3 (0.8) TPa⁻¹ (Δp = 2.23–4.90 GPa). K_NLC is calculated using the relationship K = −1/l(∂l/∂p)_T, where l is the length of the axis and (∂l/∂p)_T is the length change in pressure at constant temperature.²⁰ The value of K_NLC here is rather small compared to the massive NLC behaviour observed in the low pressure phase of Ag₃[Co(CN)₆]⁹ (K_NLC = −76(9) TPa⁻¹, Δp = 0–0.19 GPa) or the flexible MOF MIL-53(Al) (K_NLC = −28 TPa⁻¹, Δp = 0–3 GPa) for example,^17b and is much more comparable to the dense Zn formate MOF [NH₄][Zn(HCOO)₃] (−1.8(8) TPa⁻¹ (Δp = 0–0.94 GPa)).²¹ Because of the quality of the data, the exact nature, or reason for the NLC in CuPyr-II is unknown, although we aim to investigate this in the future.On increasing pressure using MeOH, the JT axis was again supressed on compression, with the Cu–N1 bond reducing in length by 0.288 Å (12%) between 0.34 and 2.15 GPa, while the Cu–O1 bond length increased by 0.216 Å (11%). The pressure at which Cu–N1 and Cu–O1 became equidistant was 1.28 GPa, measuring 2.140(5) Å and 2.131(6) Å respectively (Fig. 3 and Table S6^†). Across the entire pressure range, little to no compression or expansion was observed in the Cu–O2 bond in the 1-position of the dione in CuPyr-I, the same trend observed when compressed in FC-70. The JT switching pressure in MeOH however was 0.71 GPa higher than observed by direct compression in FC-70 (0.57 GPa). This, to our knowledge, is the first time that pressure induced JT switching has been observed to be hydrostatic media dependent.Changes to the Cu–N and Cu–O bond lengths were supported by high-pressure Raman spectroscopy of CuPyr-I, using MeOH as the hydrostatic medium (Fig. S10^†). Gradual growth of a shoulder on a band at ∼700 cm⁻¹ during compression is tentatively assigned to the Cu²⁺ coordination environment shifting from elongated to compressed JT distorted geometry. The shouldered peak becomes split above 2 GPa, after which the isosymmetric phase transition occurs.The gradual JT switch is thought to be principally responsible for reversible piezochromism in single crystals of CuPyr-I, which change in colour from green to dark red under applied pressure (Fig. 4b, S1 and S2^†). UV-visible spectroscopy confirms a bariometric blue-shift in the absorption peak at ∼700 nm assigned to d–d electronic transitions, and a red-shift of the tentatively assigned ligand-to-metal charge-transfer (LMCT) edge around 450 nm during the elongated to compressed switch (Fig. 4a and S8^†), accounting for this colour change. The red-shift is observed during compression in both Fluorinert® FC-70 and MeOH hydrostatic media, with a slightly suppressed shift measured in the latter due to filling of the framework pores (Table S2^†). Geometric switching at the metal centre leads to electronic stabilisation of the Cu²⁺ ion, as electrons transfer from higher energy d_x²−y² (Cu–O) orbitals to the lower energy d_z² (Cu–N) state (Fig. S7^†), evidenced by the blue-shift of the d–d intraconfigurational band as the d_z² (Cu–N) is progressively mixing with d_x²−y² (Cu–O) increasing its energy with respect to the lower energy d_xy and d_xz,yz levels becoming the highest energy level at the nearly compressed rhombic geometry. On the other hand, the redshift in the hesitantly assigned O²⁻ to Cu²⁺ LMCT band below 450 nm is ascribed to increase of the Cu–O bond distance and a likely bandwidth broadening with pressure both yielding a pressure redshift of the absorption band gap edge (Fig. 4a).Open in a separate windowFig. 4(a) UV-visible spectroscopy of CuPyr-I during compression in Fluorinert® FC-70 showing a gradual BLUE-shift in the d–d intraconfigurational band (∼700 nm) and a gradual red-shift of the absorption band assigned to LMCT (∼450 nm) with increasing pressure. (b) Gradual pressure-induced Jahn–Teller switch of the Cu²⁺ octahedral coordination environment in CuPyr-1 from tetragonal elongated (left, green) to rhombic compressed (right, red), causing piezochromism. Atom colouring follows previous figures.Compression of the coordination bonds was not the only distortion to take place in CuPyr-I, with the Cu-octahedra also twisting with respect to the 1-(4-pyridyl)butane-1,3-dione linkers on increasing pressure. Twisting of the Cu-octahedra in CuPyr-I with respect to the dione section of the linker could be quantified by measuring both the ∠N1Cu1O2C4 and the ∠N1Cu1O1C2 torsion angles from the X-ray data, which in MeOH gradually decrease and increase by 12.2° and 7.3°, respectively, to 2.15 GPa (Table S8^†). In FC-70, ∠N1Cu1O2C4 and ∠N1Cu1O1C2 decrease and increase by 5.4° and 2.8°, respectively, to 1.56 GPa. On increasing pressure to 1.57 GPa in a hydrostatic medium of MeOH, a difference of ∼5° for both angles was observed compared to FC-70 at 1.56 GPa. Twisting about the octahedra allows compression of the channels to take place in a ‘screw’ like fashion and has been observed in other porous materials with channel structures.²² The overall effect is to reduce the pore volume, and decrease the size of the channels (Tables S2 and S3^†). Using MeOH as a hydrostatic medium therefore appears to reduce this effect by decreasing the compressibility of the framework.It was not possible to determine the pressure dependence in other longer-chain alcohols, including ethanol (EtOH) and isopropanol (IPA), due to cracking of the crystal upon loading into the diamond anvil cell (Fig. S1^†). We believe this is a result of these longer chain alcohols acting as reducing agents, as indicated by the loss in colour of the crystals.To ascertain the origin of the hydrostatic media-induced change in the JT switching pressure and unit cell compressibility, the pore size and content were monitored as a function of pressure. A dried crystal of CuPyr-I was collected at ambient pressure and temperature in order to compare to the high-pressure data and is included in the ESI.^† The pore volume and electron density were estimated and modelled respectively using the SQUEEZE algorithm within PLATON (Tables S1–S3^†).²³ CuPyr-I under ambient pressure conditions has three symmetry equivalent channels per unit cell with a total volume of ∼1152 ų containing diethyl ether (2.5 wt%) trapped in the pores during the synthesis of the framework, confirmed by TGA analysis (Fig. S6^†).On surrounding the crystal with FC-70, direct compression of the framework occurred. The pore content remained almost constant during compression up to 0.88 GPa, inferring no change in the pore contents. On increasing pressure further to 1.56 GPa, an increase in the calculated electron density was observed (23%), though the data here were of depreciating quality and less reliable. During compression of CuPyr-I in MeOH to 0.52 GPa, the pore volume and electron density in the channels increased by 4.5% and 54%, respectively, reflecting ingress of MeOH into the pores. The electron density in the channels continued to increase to a maximum of 0.466e⁻ A⁻³ by 0.96 GPa, although the pore volume began to decrease at this pressure. The uptake of MeOH into the pores therefore results in the marked decrease in compressibility, as noted above.Previous high-pressure experiments on porous MOFs have resulted in similar behaviour on application of pressure, with the uptake of the media significantly decreasing the compressibility of the framework.²⁴ However, using different hydrostatic media to control the JT switch in any material is, to the best of our knowledge, previously unreported. On increasing pressure above 0.96 GPa, the electron density in the pores decreases, and coincides with a steady reduction in volume of the unit cell. Both an initial increase and then subsequent decrease in uptake of hydrostatic media is common in high-pressure studies of MOFs, and has been seen several times, for example in HKUST-1 (ref. 24c) and MOF-5.^24a The ingress of MeOH into the pores on initially increasing pressure to 0.52 GPa is also reflected in a twisting of the octahedra, in-particular the ∠N1Cu1O2C4 angle decreases by 5.8° in MeOH, whereas on compression in FC-70, little to no change is observed in the ∠N1Cu1O2C4 angle to 1.56 GPa. These angles represent a twisting of the dione backbone, which we speculate must interact with the MeOH molecules which penetrate into the framework.Upon compression in n-pentane, the lightest alkane that is a liquid at ambient temperature, we see different behaviour to that in MeOH or FC-70. Poor data quality permitted only the extraction of unit cell parameters but from this it can be seen that CuPyr-I has undergone the transition to CuPyr-II by 0.77 GPa. This is a significantly lower pressure than is required to induce the phase transition in MeOH (ca. 2.15 GPa). We speculate that this difference in pressure is caused by the n-pentane entering the channels at a lower pressure than MeOH due to the hydrophobic nature of the channels. This can be overcome by MeOH but not until substantially higher pressures, as seen in other MOFs that contain hydrophobic pores.²⁵On undergoing the transition to CuPyr-II at 2.78 GPa the unit cell volume quadruples, resulting in three symmetry independent channels (12 per unit cell), with the % pore volume continuing to decrease (Table S4^†). Additionally, the reflections become much broader, significantly depreciating the data quality. Nevertheless, changes in metal–ligand bond lengths and general packing features can be extracted. In particular, the transition to CuPyr-II results in two independent Cu-centres, with six independent Cu–N/O bond distances per Cu. Each exhibits a continuation of the trend seen in CuPyr-I, with the Cu–O bonds (equivalent to the Cu–O1 bond in CuPyr-I) remaining longer than the JT suppressed Cu–N bonds. However, the transition to CuPyr-II results in both an increase and decrease in three of the four Cu–N and Cu–O bonds respectively, compared to CuPyr-I at 2.15 GPa (Table S6^†). The net result is a framework which contains a Cu-centre where the coordination bonds are more equidistant, while the JT axis becomes much more prominent in the other Cu-centre, with the Cu–O dione bond continuing to increase in length. The data for CuPyr-II depreciates rapidly after the phase transition, and more work would be required to study the effect of the anisotropic compression of the JT axis in CuPyr-II on increasing pressure further.It is difficult to determine the mechanism behind the NLC behaviour observed upon compression of CuPyr-II because the phase transition results in a significant reduction in data quality. Further work will be carried out computationally in order to elucidate the structural mechanism that gives rise to the PLC followed by NLC. However, we propose this effect is inherent to this framework and the ingress of MeOH molecules into the channels allows the retention of crystallinity to allow this behaviour to be observed crystallographically.In order to determine whether the JT switch could be induced by decreasing temperature and remove any effect the ingress of hydrostatic media has into the pores on the JT switch, variable temperature X-ray diffraction measurements were undertaken on a powder and single-crystal sample. On cooling below 175 K and 150 K in a powder and single-crystal sample respectively, a phase transition was observed, however, this was to a completely different triclinic phase, hereafter referred to as CuPyr-III. The transition here appears to occur when the disordered diethyl ether becomes ordered in the pores, confirmed by determination of the structure by single-crystal X-ray diffraction, where the diethylether could be modelled inside the pore-channel (see ESI Sections 7 & 8 for details^†).In conclusion, we have presented a compression study on the newly synthesised Cu-based porous framework bis[1-(4-pyridyl)butane-1,3-dione]copper(ii), referred to as CuPyr, compressed in FC-70 to 1.56 GPa and MeOH to 4.90 GPa. In both FC-70 and MeOH hydrostatic media, the JT axis, which extends along the Cu–N pyridyl bond, steadily compresses and then switches to lie along one of the Cu–O dione bonds. Compression in MeOH results in ingress of the medium into the framework pores, which increases the JT switching pressure to 1.47 GPa, compared with 0.64 GPa during compression in Fluorinert® FC-70. Interaction of stored MeOH with the host framework prompts twisting of the ligand backbone, which is not observed in the absence of adsorbed guest. Suppression of the JT axis is accompanied by a piezochromic colour change in the single crystals from green to dark red, as confirmed by crystallographic and spectroscopic measurements. Increasing the applied pressure to at least 2.15 GPa causes the framework to undergo an isosymmetric phase transition to a previously unobserved phase, characterised by a doubling of the a/b axes. Between 2.15 GPa and 4.90 GPa, NLC behaviour is observed.This is to the best of our knowledge the first time a phase transition, NLC, piezochromic and pressure induced JT switching behaviour have been observed within the same material. We have also reported for the first time a pressure induced JT axis switch which is hydrostatic media dependent. In further analysis of this system, we intend to study the magnetic properties under ambient and high pressure.     

An engineered biosynthetic–synthetic platform for production of halogenated indolmycin antibiotics

Indolmycin is an antibiotic from Streptomyces griseus ATCC 12648 with activity against Helicobacter pylori, Plasmodium falciparum, and methicillin-resistant Staphylococcus aureus. Here we describe the use of the indolmycin biosynthetic genes in E. coli to make indolmycenic acid, a chiral intermediate in indolmycin biosynthesis, which can then be converted to indolmycin through a three-step synthesis. To expand indolmycin structural diversity, we introduce a promiscuous tryptophanyl-tRNA synthetase gene (trpS) into our E. coli production system and feed halogenated indoles to generate the corresponding indolmycenic acids, ultimately allowing us to access indolmycin derivatives through synthesis. Bioactivity testing against methicillin-resistant Staphylococcus aureus showed modest antibiotic activity for 5-, 6-, and 7-fluoro-indolmycin.

A semi-synthetic system for producing indolmycin, an antibiotic, was developed and used to make indole-substituted, halogenated derivatives of indolmycin, some with modest bioactivity against methicillin-resistant Staphylococcus aureus.

Antibiotic-resistant bacteria pose a great threat to human health,^1–4 and the rates of new antibiotic discoveries and clinical approvals have been in a steep decline since the 1980s.¹ Without the discovery and development of new antibiotics, drug-resistant pathogens, such as methicillin-resistant Staphylococcus aureus (MRSA), will become increasingly prevalent.^2–4 One strategy to increase antibiotic development has been to “rediscover” known, but underdeveloped, antibiotics.¹ One such example is indolmycin, which was originally discovered in 1960 from Streptomyces griseus ATCC 12648⁵ but was not originally developed for clinical use because of its narrow spectrum of activity^6–10 and its interference with tryptophan catabolism in the liver.^9,10 However, reignited interest in this old antibiotic led to the discovery of its activity against Helicobacter pylori,⁶Plasmodium falciparum,¹¹ and MRSA.¹² For MRSA, indolmycin was found to be active against mupirocin- and fuscidic acid-resistant MRSA strains, with strains resistant to indolmycin emerging infrequently and with reduced fitness compared to sensitive strains.¹² In addition, indolmycin has been shown to have minimal activity against common members of the human microbiota, suggesting that its narrow spectrum of activity is an asset.⁶ The first indole-substituted derivatives, 5-hydroxy and 5-methoxyindolmycin, were made by precursor-directed feeding of the indolmycin producer, Streptomyces griseus ATCC 12648, and they showed modest improvements in bioactivity against S. aureus and Escherichia coli.¹³ Two practical synthetic routes to indolmycin and some indole-substituted derivatives have been reported more recently,^14,15 which enabled access to a small variety of indole-substituted derivatives. Additionally, a previous patent has described synthetic methods to produce a variety of derivatives; however, these methods do not appear to offer stereochemical control, and some require tailoring steps specific to each analog.¹⁶ Therefore, further development of indolmycin would benefit from a simpler diversification method that could be applied to produce a wider variety of analogs with stereochemical control.Inspired by early biosynthetic studies,^17,18 our group previously identified the indolmycin gene cluster and elucidated the biosynthetic pathway, demonstrating that indolmycin (1) is assembled from tryptophan, arginine and S-adenosylmethionine (SAM) in a three-part process (Fig. 1).¹⁹ In the first part, l-arginine is oxidized by Ind4 in an oxygen- and PLP-dependent reaction to 4,5-dehydro-2-iminoarginine, which is then enantioselectively reduced by imine reductase Ind5 and its chaperone Ind6 to 4,5-dehydro-d-arginine. In the second part, tryptophan (2) is deaminated by PLP-dependent transaminases, giving indole pyruvate (3). Compound 3 is then methylated by SAM-dependent C-methyltransferase Ind1 to 3-methyl-indolepyruvate (4) which is reduced by NADH-dependent ketone reductase Ind2 to form indolmycenic acid (5). Then, in the third part, 4,5-dehydro-d-arginine and 5 are coupled in an ATP-dependent fashion by Ind3 and Ind6, resulting in an oxazolinone-cyclized molecule, N-desmethyl-indolmycin, which is finally N-methylated by Ind7, a SAM-dependent N-methyltransferase, to form 1.Open in a separate windowFig. 1Indolmycin biosynthesis from Streptomyces griseus ATCC 12648. (a) Indolmycin biosynthetic gene cluster. (b) Indolmycin biosynthetic pathway from Streptomyces griseus ATCC 12648. (c) Semi-synthetic scheme towards indolmycin and derivatives using indolmycin biosynthetic genes. The dashed arrow indicates a predicted side-product based on LC-MS analysis.Armed with the elucidated biosynthetic pathway for 1, we set out to create an in vivo system to make 1 in E. coli. We first cloned all necessary genes into four plasmids and co-expressed the genes in E. coli (Fig. S1^†), a strain which we named E. coli I1234670P5 (Table S1^†). We found that the genes needed to produce indolmycin in E. coli were ind1, ind2, ind3, ind4, ind6, ind7, ind0 and pel5, a homologous gene of ind5 from Paenibacillus elgii B69 showing better production of active protein in E. coli.^20,21 We also relied on the activity of endogenous E. coli aminotransferases to catalyze the initial tryptophan deamination step. However, only a small amount of 1 was produced (∼170 μg L⁻¹ of bacterial culture) and the yield could not be improved despite our best efforts (Fig. 2a). However, we found that this construct produced substantial amounts of 5 ([M + H]⁺ = 220 m/z) at 40–50 mg L⁻¹ of culture, along with a shunt product, C-desmethyl-indolmycenic acid (6; [M + H]⁺ = 206 m/z).Open in a separate windowFig. 2Biosynthetic production of 5 and semi-synthetic production of 1. (a) Extracted ion chromatograms show production of 5 with minimal production of 1 from E. coli I1234670P5. (b) Synthetic scheme to 1 from 5, adapted from literature methods.^14,15 (c) Total ion chromatogram of compound 1 isolated after semi-synthesis and final purification by semi-preparative HPLC. Compounds are indicated with coloured boxes and numbered.Compound 5 itself has been a focus of total synthetic efforts toward 1, as it is the key chiral precursor.^{14,15,22–26} Since production of 5 was much higher than that of 1 from E. coli I1234670P5, we pursued a semi-synthetic method of obtaining 1 using our biosynthetic platform to access 5, combined with a three-step chemical transformation (Fig. 2b). We attempted to remove extraneous genes from our biosynthetic platform by only including ind0, ind1, and ind2, but the changes resulted in reduced amounts of 5 (Fig. S2^†). At this time, it is unclear which of the other genes may be contributing to the production of compound 5. Therefore, we employed the full eight-gene construct toward synthesis of 5. We then adapted the three-step synthesis to make indolmycin,^14,15 in which purified 5 was esterified to make the ethyl ester (7; [M + H]⁺ = 248 m/z; Fig. S3a^†), cyclized to give N-desmethyl-indolmycin (8; [M + H]⁺ = 244 m/z; Fig. S3b and S4a^†), and methylated at the exocyclic nitrogen to give 1 ([M + H]⁺ = 258 m/z; Fig. 2, S4b and Table S4^†).Then, we wanted to make derivatives of 1 from indole derivatives, which are more widely accessible than derivatives of 2. The tryptophan synthase (TrpS) from Salmonella enterica has been previously shown to couple a wide variety of indole derivatives to l-serine to generate derivatives of 2.²⁷ We were able to replace pel5 with trpS in our biosynthetic platform without a reduction in the amount of 5 produced (Fig. S1b^†), and we named the resulting strain E. coli I1234670TS. When we fed 5-fluoroindole to E. coli I1234670TS, we observed increased production of fluorinated metabolites, 5-fluoro-indolmycenic acid (5F-5; [M + H]⁺ = 238 m/z) and 5-fluoro-C-desmethyl-indolmycenic acid (5F-6; [M + H]⁺ = 224 m/z) (Fig. 3a). We then optimized the feeding conditions, finding that 5F-5 amounts were optimal when we fed E. coli I1234670TS with 0.5 mM 5-fluoroindole per day over two days (Fig. S5^†).Open in a separate windowFig. 3Addition of the trpS gene to the biosynthetic platform allows incorporation of substituted indoles into 5. (a) LC-MS analysis of strains with and without trpS (E. coli I1234670TS and E. coli I1234670P5, respectively) when fed 5-fluoroindole. Chemical structures of compounds with the corresponding extracted ion chromatogram are shown to the right of the traces. (b) Indole derivatives tested. Dark blue shows indoles that were incorporated into an analog of 5 (>48% of underivatized 5 by LC-MS analysis; Table S5^†) and further purified; medium blue indicates indoles that were incorporated into an analog of 5 at lower levels (6–22% of underivatized 5 by LC-MS analysis; Table S5^†) but were not further verified by purification; and light blue indicates indoles that did not show detectable incorporation into 5.To determine the scope of indole derivatives accepted by our biosynthetic platform, we fed a variety of indoles to E. coli I1234670TS and monitored the production of 5 and its derivatives by LC-MS. Out of the indoles tested, we found that fluorinated and chlorinated indoles substituted at the 5-, 6- and 7-positions were the best accepted by the biosynthetic platform (Fig. 3b, S6 and S7^†). We predict that lower acceptance of indoles substituted at the 4-position may be due to steric hindrance, as 4-fluoroindole was moderately accepted, while 4-chloroindole was not observed at all. Although we observed LC-MS peaks consistent with conversion of some of the azaindoles and hydroxyindoles into 5 derivatives, further work is required to confirm, optimize and scale up the purification of these compounds (Fig. S7^†). Cultures producing derivatives of 5, substituted at 5-, 6- and 7-positions, were further scaled up for purification of the 5-derivatives and downstream synthesis of 1-derivatives (Fig. S8–S10 and Table S4^†). Each purified derivative of 5 and 1 was characterized by HR-MS and NMR (Table S3 and ESI Methods^†). Overall, cultures fed with the fluorinated indoles produced a higher amount of 5-derivatives than the cultures fed with the chlorinated indoles. 1 and its derivatives were tested against MRSA (Fig. S11^†). While the fluorinated derivatives showed bioactivity, the chlorinated derivatives of 1 did not show bioactivity at the maximum amount tested in the disk diffusion assay (30 μg). We determined MIC₅₀ values for each fluorinated compound (Table 1). The MIC₅₀ values demonstrate that 1 is a more potent inhibitor of MRSA than its derivatives, while 6F-1 showed the most potent inhibition of MRSA compared to any of the derivatives, followed by 7F-1 and 5F-1. The lack of bioactivity of the chlorinated compounds may be due to the bulky chlorinated substituent hindering the compounds'' abilities to bind to the tryptophanyl-tRNA synthetase (TrpRS) target, which is supported by docking studies of the analogs into a bacterial TrpRS structure (Fig. S12^†).MIC₅₀ values determined for 1 and its derivatives against MRSA. Values represent the average of three replicates ± the standard deviation. For 1, 5F-1, 6F-1 and 7F-1, concentrations ranging from 50 μg mL⁻¹ to 0.128 ng mL⁻¹ were tested, and for 5Cl-1, 6Cl-1 and 7Cl-1, concentrations ranging from 200 μg mL⁻¹ to 0.512 ng mL⁻¹ were tested. The reported MIC₅₀ value for indolmycin is 0.5 μg mL⁻¹ (range: 0.125–2 μg mL⁻¹) against MRSA¹²

Inhibitors of thiol-mediated uptake

Ellman''s reagent has caused substantial confusion and concern as a probe for thiol-mediated uptake because it is the only established inhibitor available but works neither efficiently nor reliably. Here we use fluorescent cyclic oligochalcogenides that enter cells by thiol-mediated uptake to systematically screen for more potent inhibitors, including epidithiodiketopiperazines, benzopolysulfanes, disulfide-bridged γ-turned peptides, heteroaromatic sulfones and cyclic thiosulfonates, thiosulfinates and disulfides. With nanomolar activity, the best inhibitors identified are more than 5000 times better than Ellman''s reagent. Different activities found with different reporters reveal thiol-mediated uptake as a complex multitarget process. Preliminary results on the inhibition of the cellular uptake of pseudo-lentivectors expressing SARS-CoV-2 spike protein do not exclude potential of efficient inhibitors of thiol-mediated uptake for the development of new antivirals.

Thiol-reactive inhibitors for the cellular entry of cyclic oligochalcogenide (COC) transporters and SARS-CoV-2 spike pseudo-lentivirus are reported.

Thiol-mediated uptake^1–10 has been developed to explain surprisingly efficient cellular uptake of substrates attached to thiol-reactive groups, most notably disulfides. The key step of this mechanism is the dynamic covalent thiol-disulfide exchange between disulfides of the substrates and exofacial thiols on cell surfaces (Fig. 1). The covalently bound substrate then enters the cell either by fusion, endocytosis, or direct translocation across the plasma membrane into the cytosol. Thiol-disulfide exchange has been confirmed to play an essential role in the cellular entry of some viruses^1,11–14 and toxins.² Indeed, diphtheria toxin and HIV were among the first to be recognized to enter cells via thiol-mediated uptake.^1,2 The involvement of cell-surface thiols in cellular uptake is most often probed by inhibition with Ellman''s reagent (DTNB). However, this test is not always reliable, in part due to the comparably poor reactivity of DTNB, and the comparably high reactivity of the disulfide obtained as a product. Thus, the importance of thiol-mediated uptake for viral entry and beyond remains, at least in part, unclear.Open in a separate windowFig. 1In thiol-mediated uptake, dynamic covalent exchange with thiols on the cell surface precedes entry through different mechanisms. Inhibition of thiol-mediated uptake by removal of exofacial thiols and disulfides could thus afford new antivirals.We became interested in thiol-mediated uptake^3–5 while studying the cytosolic delivery of substrates such as drugs, probes and also larger objects like proteins or quantum dots with cell-penetrating poly(disulfide)s.⁶ Our recent focus shifted to cyclic oligochalcogenides (COCs) to increase speed and selectivity of dynamic covalent thiol-oligochalcogenide exchange, and, most importantly, to assure reversibility, i.e., mobility during uptake, with a covalently tethered, intramolecular leaving group.⁷ With increasingly unorthodox COC chemistry, from strained disulfides^7,8 and diselenides⁹ to adaptive dynamic covalent networks produced by polysulfanes,¹⁰ uptake activities steadily increased. Their high activities suggested that the same, or complementary, COCs could also function as powerful inhibitors of thiol-mediated uptake that ultimately might perhaps lead to antivirals. In the following, this hypothesis is developed further.Fluorescently labeled COCs 1⁸ and 2¹⁰ were selected as reporters for the screening of thiol-mediated uptake inhibitors because of their high activity, their destination in the cytosol, and their different characteristics (Fig. 2). The COC in 1 is an epidithiodiketopiperazine (ETP). With a CSSC dihedral angle ∼0°, ETPs drive ring tension to the extreme.^15,16 Ring-opening thiol-disulfide exchange is ultrafast, and the released thiols are acidic enough to continue exchanging in neutral water, including ring closure.⁸ This unique exchange chemistry coincides with efficient cellular uptake and poor retention on thiol affinity columns.⁸Open in a separate windowFig. 2Structure of reporters 1 and 2 and inhibitor candidates 3–30 with their concentrations needed to inhibit by ∼15% (MIC) the uptake of 1 (1 h pre-incubation with inhibitors, 30 min incubation with reporter, filled symbols) and 2 (4 h pre-incubation, empty symbols). Red squares: ETPs; orange circles: BPSs; blue upward triangles: heteroaromatic sulfones; purple diamonds: thiosulfonates; magenta downward triangles: di- and polysulfides; brown hexagons: thiosulfinates. Symbols with upward arrows: MIC not reached at the highest concentration tested. Symbols with downward arrows indicate the lowest concentration tested already exceeds the MIC. (a) Similarly active upon co-incubation of reporters and inhibitor; (b–d) similarly (b), less (c), or more (d) active upon co-incubation in the presence of serum (mostly 6 h); (e) pre-incubation for 15 min; (f) isomerizes into cis22; (g) V-shaped DRC (see Fig. 3f); (h) pre-incubation for 30 min, co-incubation with 2; (i) mixture of regioisomers.The COC in 2 is a benzopolysulfane (BPS). Like ETPs, BPSs occur in natural products and have inspired total synthesis.¹⁷ Unlike ETPs, BPSs are not strained but evolve into adaptive networks of extreme sulfur species for cells to select from. Uptake efficiencies and retention on thiol affinity columns exceed other COCs clearly.^10,18With COCs 1 and 2 as cell-penetrating reporters, a fully automated, fluorescent microscopy image-based high-content high-throughput (HCHT)¹⁹ inhibitor screening assay was developed. HeLa cells in multiwell plates are incubated with a reporter at constant and inhibitors at varying concentrations and incubation times. Hindered reporter uptake then causes decrease of fluorescence inside of cells (Fig. 3a). Automated data analysis¹⁹ was established to extract average fluorescence intensity per cell and, at the same time, cell viability from propidium iodide negative nuclei count (Fig. 3 and S3–S6^†). Standard assay conditions consisted of pre-incubation of HeLa cells with inhibitors for different periods of time, followed by the removal of inhibitors and the addition of reporters, thus excluding possible interactions between the two in the extracellular environment. In alternative co-incubation conditions, inhibitors were not removed before the addition of reporters to allow for eventual interactions between the two.Open in a separate windowFig. 3(a) Fluorescence image of HCHT plates (4 images per well) with HeLa cells pre-incubated with 6 (30 min) followed by co-incubation with 1 (left) and 2 (right, 10 μM each) for constant 30 min. (b–f) HCHT data showing relative fluorescence intensity (filled symbols) and cell viability (empty symbols) of HeLa cells after (b) pre-incubation with 4 for 1 h, followed by washing and incubation with 1 (top), or pre-incubation with 4 for 30 min, followed by co-incubation with 4 and 2 (bottom). (c) As in (b) with 18. (d) As in (b) after incubation for 4 h with 16 followed by incubation with 2. (e) As in (b) after pre-incubation with 11 (circles), 14 (crosses), or 21 (diamonds) for 15 min, followed by washing and incubation with 1. (f) As in (b) after pre-incubation with 20 (30 min), followed by washing and incubation with 1.Among the very high number of thiol-reactive probes, compounds 3–30 were selected based on promise, experience, availability and accessibility. Main focus was on COCs offering increasingly extreme sulfur chemistry because dynamic covalent thiol-oligochalcogenide exchange with different intramolecular leaving groups promises access to different exchange cascades for the intramolecular and, perhaps, also intermolecular crosslinking of the target proteins. More hydrophilic, often anionic COCs were preferred to prevent diffusion into cells and thus minimize toxicity. The expectation was that from such a sketchy outline of an immense chemical space, leads could be identified for future, more systematic exploration. Reporters 1 and 2 and candidates 3–30 were prepared by substantial multistep synthesis (Schemes S1–S11 and Fig. S47–S93,^† commercially available: 20, 25, 30). Inhibitors were numbered in the order of efficiency against reporter 1, evaluated by their minimum inhibitory concentrations (MICs), i.e., concentrations that cause a ∼15% reduction of reporter uptake in cells (Fig. 2 and Tables S1–S37^†). We chose to use MICs because half-maximal inhibitions could not always be reached due to the onset of toxicity, formally anticooperative, or even V-shaped dose–response curves (DRCs, e.g., Fig. 3b–f, all DRCs can be found in the ESI, Fig. S7–S43^†). MICs are usually below the half-maximal cell growth inhibition concentration (GI₅₀, Tables S1–S37^†).Among the most potent inhibitors of ETP reporter 1 were ETPs 4 and 5 (Fig. 2, ​,3b).3b). This intriguing self-inhibition was even surpassed by the expanded cyclic tetrasulfide ETP₄3 (MIC < 0.1 μM), which was of interest because they are much poorer transporters.¹⁰ Further formal ring expansion leads to cyclic pentasulfides BPS₅6 as equally outstanding inhibitors (MIC ≈ 0.3 μM). This trend toward the adaptive networks, reminiscent of elemental sulfur chemistry, did not extend toward inorganic polysulfides 13 (MIC ≈ 20 μM). ETPs 4 and 5 were sensitive to modification of the carboxylate, with the cationic 12 being the worst (MIC ≈ 30 μM) and the neutral glucose hemiacetal 7 the most promising (MIC ≈ 0.5 μM).Although this study focuses on increasingly extreme dynamic covalent COC chemistry, the inclusion of one example for covalent C–S bond formation was of interest for comparison. The classical iodoacetamides⁷ and maleimides⁴ were more toxic than active (not shown). However, nucleophilic aromatic substitution of heteroaromatic sulfones,²⁰ just developed for the efficient bioorthogonal conversion of thiols into sulfides, was more promising. Weaker than dynamic covalent COCs, this irreversible inhibition was best with benzoxazole 11 (MIC ≈ 15 μM) and decreased in accordance with reactivity toward free thiols to oxadiazole 14 and benzothiazole 21 (MIC ≈ 300 μM, Fig. 3e).At constant pH, Ellman''s reagent 20 was confirmed to be erratic also in this assay. The DRC showed minor inhibition up to around 2 mM, which disappeared again at higher concentrations (Fig. 3f). Other cyclic disulfides were inactive as well (28–30). Also disappointing were oxidized disulfides, that is thiosulfinates, including allicin 25, the main odorant component of garlic,^21,22 oxidized cystine 26 and oxidized lipoic acid 27. Thiosulfinates were of interest because they should selectively target the vicinal thiols of reduced disulfides bridges, producing two disulfides.²³ The most active trans dithioerythrol (DTE) thiosulfinate 17 isomerized with time into the less active, hydrogen-bonded cis isomer 22 (Fig. S46^†).Reporter 2 was more difficult to inhibit than 1, as expected from high activity with extreme retention on thiol affinity columns.^10,18 For instance, BPS 6 was very efficient against ETP 1 but much less active against BPS 2 (Fig. 3a), although longer pre-incubation could lower the MIC down to 4 μM (Fig. 2, S41^†). The complementary ETP 4 “self-inhibited” ETP 1 but was also unable to inhibit BPS 2 as efficiently (Fig. 3b). Among the best inhibitors of BPS 2 upon co-incubation were disulfide bridged γ-turn²⁴ peptides 18 and 19 (MIC ≈ 5 μM), both less active against 1 (MIC ≈ 300 μM, Fig. 3c). Disulfide-bridged γ-turn CXC peptides consist of an 11-membered ring with significant Prelog strain. They were introduced by Wu and coworkers as transporters for efficient cytosolic delivery.⁵ The cyclic thiosulfonates 15 and 16 showed promising activities against both 1 and 2, and were tolerant toward the presence of serum (Fig. 2d, S33 and S42^†). Contrary to thiosulfinate 27, the oxidation of lipoic acid to pure thiosulfonates was not successful so far. However, weakly detectable activity of the lipoyl-glutamate conjugate oxidized to the thiosulfinate (MIC ≈ 350 μM, not shown) compared to the inactive thiosulfinate 27 implied that lipoic acid oxidized to the thiosulfonate would also be less active than the glutamate conjugate 15.The oxidized DTE 16^25–28 was particularly intriguing because it was more potent against 2 and could achieve nearly complete inhibition (MIC ∼ 20 μM, Fig. 3d). Highly selective for thiols, the cyclic thiosulfonate 16 was stable for weeks at room temperature, without precaution, in all solvents tested. The disulfides and sulfinates obtained from exchange with thiols were stable as well, and the latter can further react with disulfides²⁷ for intramolecular or eventually intermolecular crosslinking of the target proteins.The overall mismatched inhibition profiles found for reporters 1 and 2 supported that thiol-mediated uptake proceeds through a series of at least partially uncoupled parallel multitarget systems instead of a specific single protein or membrane target. From proteomics studies with cysteine-reactive irreversible probes, it is known that different probes generally target different proteins.^29b Proteomics analysis^29a for asparagusic acid derived transporters supports the involvement of many targets beyond the commonly considered protein disulfide isomerases and the confirmed transferrin receptor.^{12–14,26–30} The unusual, formally anti-cooperative (Hill coefficients < 1) DRCs further supported thiol-mediated uptake as complex multitarget systems.Despite the complexity of these systems, results did not much depend on assay conditions. Compared to the standard protocol of pre-incubation with inhibitors followed by inhibitor removal and incubation with reporters 1 or 2 for detection, the co-incubation protocol, in which pre-incubation with inhibitors is followed by co-incubation with reporters 1 or 2 without inhibitor removal, gave reasonably similar results (Fig. 2). Inhibition characteristics naturally depended on pre-incubation time, with weaker activities at shorter and longer times, reflecting incomplete exchange and cellular response or other ways of inhibitor destruction, respectively. The presence of serum also did not affect the activities much (Fig. 2b–d).Preliminary studies on antiviral activity were performed with pseudo-lentivectors³¹ that express the D614G mutant¹¹ of the SARS-CoV-2 spike protein and code for a luciferase reporter gene, which is expressed by the infected cells.¹² A549 human lung alveolar basal epithelium cell line constitutively overexpressing ACE2 and TMPRSS2 was selected to facilitate the entry of the SARS-CoV-2 spike pseudo-lentivirus. The most significant activities were found for DTE thiosulfonate 16 with an IC₅₀ around 50 μM, while toxicity was detected only at 500 μM (Fig. S44^†). The onset of inhibition could be observed for tetrasulfide ETP 3 at 50 μM, but it coincided with the appearance of cytotoxicity. Protease inhibition is less likely to be the mode of action, as similar activity was found with wild type A549 cells transduced with a standard lentivirus expressing vesicular-stomatitis virus G surface protein VSVG (Fig. S45^†).¹³ Short incubation times of cells and inhibitors before the addition of viruses disfavored contributions from changes in gene expression. More detailed studies are ongoing.The lessons learned from this study are that, firstly, thiol-mediated uptake can be inhibited efficiently by thiol-reactive reagents, confirming that thiol-mediated uptake exists and transporters like ETP 1 and BPS 2 do not simply diffuse into cells; the best inhibitors are more than 5000 times better than Ellman''s reagent. Secondly, inhibitor efficiencies vary with the transporters, supporting that thiol-mediated uptake operates as a complex multitarget system. The best inhibitors are COCs that operate with fast dynamic covalent exchange, suggesting that the reversibility provided by COCs is important. The inhibition of thiol-mediated uptake might contribute to activities of thiol-reactive antivirals such as 16, ETPs or ebselen, although they have been shown to bind to zinc fingers or inhibit proteases.^{16,25,32–34} Finally, the inhibitors reported here could also be of interest for delivery applications and might be worth investigation with regard to antiviral activity. We currently plan to focus more systematically on the most promising leads within COCs, particularly cyclic thiosulfonates, and to expand the screening campaign toward new attractive motifs.^33–35     

Nitric oxide monooxygenation (NOM) reaction of cobalt-nitrosyl {Co(NO)}8 to CoII-nitrito {CoII(NO2−)}: base induced hydrogen gas (H2) evolution

Here, we report the nitric oxide monooxygenation (NOM) reactions of a Co^III-nitrosyl complex (1, {Co-NO}⁸) in the presence of mono-oxygen reactive species, i.e., a base (OH⁻, tetrabutylammonium hydroxide (TBAOH) or NaOH/15-crown-5), an oxide (O²⁻ or Na₂O/15-crown-5) and water (H₂O). The reaction of 1 with OH⁻ produces a Co^II-nitrito complex {3, (Co^II-NO₂⁻)} and hydrogen gas (H₂), via the formation of a putative N-bound Co-nitrous acid intermediate (2, {Co-NOOH}⁺). The homolytic cleavage of the O–H bond of proposed [Co-NOOH]⁺ releases H₂via a presumed Co^III-H intermediate. In another reaction, 1 generates Co^II-NO₂⁻ when reacted with O²⁻via an expected Co^I-nitro (4) intermediate. However, complex 1 is found to be unreactive towards H₂O. Mechanistic investigations using ¹⁵N-labeled-¹⁵NO and ²H-labeled-NaO²H (NaOD) evidently revealed that the N-atom in Co^II-NO₂⁻ and the H-atom in H₂ gas are derived from the nitrosyl ligand and OH⁻ moiety, respectively.

Base-induced hydrogen (H₂) gas evolution in the nitric oxide monoxygenation reaction.

As a radical species, nitric oxide (NO) has attracted great interest from the scientific community due to its major role in various physiological processes such as neurotransmission, vascular regulation, platelet disaggregation and immune responses to multiple infections.¹ Nitric oxide synthase (NOS),² and nitrite reductase (NiR)³ enzymes are involved in the biosynthesis of NO. NOSs produce NO by the oxidation of the guanidine nitrogen in l-arginine.⁴ However, in mammals and bacteria, NO₂⁻ is reduced to NO by NiRs in the presence of protons, i.e., NO₂⁻ + e⁻ + 2H⁺ → NO + H₂O.⁵ Biological dysfunctions may cause overproduction of NO, and being radical it leads to the generation of reactive nitrogen species (RNS), i.e., peroxynitrite (PN, OONO⁻)⁶ and nitrogen dioxide (˙NO₂),⁷ upon reaction with reactive oxygen species (ROS) such as superoxide (O₂˙⁻),⁸ peroxide (H₂O₂),⁹ and dioxygen (O₂).¹⁰ Hence, it is essential to maintain an optimal level of NO. In this regard, nitric oxide dioxygenases (NODs)¹¹ are available in bio-systems to convert excess NO to biologically benign nitrate (NO₃⁻).¹²NO₂⁻ + Fe^II + H⁺ ↔ NO + Fe^III + OH⁻1[M–NO]ⁿ + 2OH⁻ → [M–NO₂]⁽ⁿ⁻²⁾ + H₂O2NOD enzymes generate NO₃⁻ from NO;^11b,12−13 however, the formation of NO₂⁻ from NO is still under investigation. Clarkson and Bosolo reported NO₂⁻ formation in the reaction of Co^III-NO and O₂.¹⁴ Nam and co-workers showed the generation of Co^II-NO₂⁻ from Co^III-NO upon reaction with O₂˙⁻.¹⁵ Recently, Mondal and co-workers reported NO₂⁻ formation in the reaction of Co^II-NO with O₂.¹⁶ Apart from cobalt, the formation of Cu^II-NO₂⁻ was also observed in the reaction of Cu^I-NO and O₂.¹⁷ For metal-dioxygen adducts, i.e., Cr^III-O₂˙⁻ and Mn^IV-O₂²⁻, NOD reactions led to the generation of Cr^III-NO₂⁻ (ref. 18) and Mn^V O + NO₂⁻,¹⁹ respectively. However, the NOD reaction of Fe^III-O₂˙⁻ and Fe^III-O₂²⁻ with NO and NO⁺, respectively, generated Fe^III-NO₃⁻via Fe^IV O and ˙NO₂.²⁰ Ford suggested that the reaction of ferric-heme nitrosyl with hydroxide leads to the formation of NO₂⁻ and H⁺.¹² Lehnert and co-workers reported heme-based Fe-nitrosyl complexes²¹ showing different chemistries due to the Fe^II-NO⁺ type electronic structures. On the other hand, Bryan proposed that the one-electron reduction of NO₂⁻ to NO in ferrous heme protein is reversible (eqn (1)).²² Also, it is proposed that excess NO in biological systems is converted to NO₂⁻ and produces one equivalent of H⁺ upon reaction with ˙OH.²³ Previously reported reactivity of M–NOs of Fe²⁴ with OH⁻ suggested the formation of NO₂⁻ and one equivalent of H⁺, where H⁺ further reacts with one equivalent of OH⁻ and produces H₂O (eqn (2)).²⁵Here in this report, we explore the mechanistic aspects of nitric oxide monooxygenation (NOM) reactions of the Co^III-nitrosyl complex, [(12TMC)Co^III(NO⁻)]²⁺/{Co(NO)}⁸ (1),^15,26 bearing the 12TMC ligand (12TMC = 1,4,7,10-tetramethyl-1,4,7,10-tetraazacyclododecane) with mono-oxygen reactive species (O²⁻, OH⁻ and H₂O) (Scheme 1). Complex 1 reacts with the base (OH⁻, tetrabutylammonium hydroxide (TBAOH)/or NaOH in the presence of 15-crown-5 as the OH⁻ source) and generates the corresponding Co^II-nitrito complex, [(12TMC)Co^II(NO₂⁻)]⁺ (3), with the evolution of hydrogen gas (H₂) via the formation of a plausible N-bound Co-nitrous acid intermediate ([Co-NOOH]⁺, 2) in CH₃CN at 273 K (Scheme 1, reaction (I)). Also, when 1 reacts with the oxide (O²⁻ or Na₂O in the presence of 15-crown-5), it generates the Co^II-nitrito complex (3) via a probable Co^I-nitro, [(12TMC)Co^I(NO₂⁻)] (4), intermediate (Scheme 1, reaction (II)); however, 1 does not react with water (Scheme 1, reaction (III)). Mechanistic investigations using ¹⁵N-labeled-¹⁵NO, D-labeled-NaOD and ¹⁸O-labelled-¹⁸OH⁻ demonstrated, unambiguously, that the N and O-atoms in the NO₂⁻ ligand of 3 resulted from NO and OH⁻ moieties; however, the H-atoms of H₂ are derived from OH⁻. To the extent of our knowledge, the present work reports the very first systematic study of Co^III-nitrosyl complex reactions with H₂O, OH⁻ and O²⁻. This new finding presents an alternative route for NO₂⁻ generation in biosystems, and also illustrates a new pathway of H₂ evolution, in addition to the reported literature.^12,27Open in a separate windowScheme 1Nitric oxide monooxygenation (NOM) reactions of cobalt-nitrosyl complex (1) in the presence of a base (OH⁻), sodium oxide (Na₂O) and water (H₂O).To further explore the chemistry of [(12TMC)Co^III(NO⁻)]²⁺ (1),^15,26 and the mechanistic insights of NOM reactions, we have reacted it with a base (OH⁻), an oxide (O²⁻), and water (H₂O). When complex 1 was reacted with TBAOH in CH₃CN, the color of complex 1 changed to light pink from dark pink. In this reaction, the characteristic absorption band of 1 (370 nm) disappears within 2 minutes (Fig. 1a; ESI, Experimental section (ES) and Fig. S1a^†), producing a Co^II-nitrito complex, [(12TMC)Co^II(NO₂⁻)]⁺ (3), with H₂ (Scheme 1, reaction (Ib)), in contrast to the previous reports on base induced NOM reactions (eqn (2)).^12,25,28 The spectral titration data confirmed that the ratio-metric equivalent of OH⁻ to 1 was 1 : 1 (ESI, Fig. S1b^†). 3 was determined to be [(12TMC)Co^II(NO₂⁻)](BF₄) based on various spectroscopic and structural characterization experiments (vide infra).^15,26bOpen in a separate windowFig. 1(a) UV-vis spectral changes of 1 (0.50 mM, black line) upon addition of OH⁻ (1 equiv.) in CH₃CN under Ar at 273 K. Black line (1) changed to red line (3) upon addition of OH⁻. Inset: IR spectra of 3-¹⁴NO₂⁻ (blue line) and 3-¹⁵NO₂⁻ (red line) in KBr. (b) ESI-MS spectra of 3. The peak at 333.2 is assigned to [(12TMC)Co^II(NO₂)]⁺ (calcd m/z 333.1). Inset: isotopic distribution pattern for 3-¹⁴NO₂⁻ (red line) and 3-¹⁵NO₂⁻ (blue line).The FT-IR spectrum of 3 showed a characteristic peak for nitrite stretching at 1271 cm⁻¹ (Co^II-14NO₂⁻) and shifted to 1245 cm⁻¹ (Co^II-15NO₂⁻) when 3 was prepared by reacting ¹⁵N-labeled NO (Co^III-15NO) with OH⁻ (Inset, Fig. 1a and Fig. S2^†). The shifting of NO₂⁻ stretching (Δ = 30 cm⁻¹) indicates that the N-atom in the NO₂⁻ ligand is derived from Co^III-15NO. The ESI-MS spectrum of 3 showed a prominent peak at m/z 333.2, [(12TMC)Co^II(¹⁴NO₂⁻)]⁺ (calcd m/z 333.2), which shifted to 334.2, [(12TMC)Co^II(¹⁵NO₂⁻)]⁺ (calcd m/z 334.2), when the reaction was performed with Co^III-15NO (Inset, Fig. 1b; ESI, Fig. S3a^†); indicating clearly that NO₂⁻ in 3 was derived from the NO moiety of 1. In addition, we have reacted 1 with Na¹⁸OH (ES and ESI^†), in order to follow the source of the second O-atom in 3-NO₂⁻. The ESI-MS spectrum of the reaction mixture, obtained by reacting 1 with Na¹⁸OH, showed a prominent peak at m/z 335.2, [(12TMC)Co^II(¹⁸ONO⁻)]⁺ (calcd m/z 335.2), (SI, Fig. S3b^†) indicating clearly that NO₂⁻ in 3 was derived from ¹⁸OH⁻. The ¹H NMR spectrum of 3 did not show any signal for aliphatic protons of the 12TMC ligand, suggesting a bivalent cobalt center (Fig. S4^†).^26b Furthermore, we have determined the magnetic moment of 3, using Evans'' method, and it was found to be 4.62 BM, suggesting a high spin Co(ii) metal center with three unpaired electrons (ESI^† and ES).²⁹ The exact conformation of 3 was provided by single-crystal X-ray crystallographic analysis (Fig. 2b, ESI, ES, Fig. S5, and Tables T1 and T2^†) and similar to that of previously reported Co^II-NO₂⁻/M^II-NO₂⁻.^15,26b Also, we have quantified the amount of nitrite (90 ± 5%), formed in the above reaction, using the Griess reagent (ESI, ES, and Fig. S6^†).Open in a separate windowFig. 2Displacement ellipsoid plot (20% probability) of 3 at 100 K. Disordered C-atoms of the TMC ring, anion and H-atoms have been removed for clarity.As is known from the literature, a metal-nitrous acid intermediate may form either by the reaction of a metal-nitrosyl with a base²⁷ or by the metal-nitrite reaction with an acid (nitrite reduction chemistry);^26b however, the products of both the reactions are different. Here, for the first time, we have explored the reaction of Co^III-nitrosyl (1) with a base. In this reaction, it is clear that the formation of Co^II-nitrito would be accomplished by the release of H₂ gas via the generation of a transient N-bound [Co-(NOOH)]⁺ intermediate (Scheme 2, reaction (II)). The formation of Co^II-NO₂⁻ (3) from the [Co-(NOOH)]⁺ intermediate is likely to proceed by either (i) homolytic cleavage of the O–H bond and release of H₂via the proposed Co^III-H transient species (Co^III-H = Co^II + 1/2H₂)³⁰ (Scheme 2, reaction (III)), as reported in previous literature where the reduced cobalt, in a number of different ligand environments, is a good H⁺ reduction catalyst and generates H₂ gas via a Co^III-H intermediate³¹ or (ii) heterolytic cleavage of the O–H bond and the formation of Co^I-NO₂⁻ + H⁺.²⁷ In the present study, we observed the formation of 3 and H₂via the plausible homolytic cleavage of the NOO–H moiety of 2 as shown in Scheme 2, in contrast to the previous reports on base-induced reactions on metal-nitrosyls (eqn (3)).²⁷ Taking together both possibilities, (i) is the most reasonable pathway for the NOM reaction of complex 1 in the presence of a base (as shown in Scheme 2, reaction (III)). And the reaction is believed to go through a Co^III-H intermediate as reported previously in Co^I-induced H⁺ reduction in different ligand frameworks and based on literature precedence, we believe that complex 1 acts in a similar manner.³¹Open in a separate windowScheme 2NOM reaction of complex 1 in the presence of OH⁻, showing the generation of Co^II-nitrito (3) and H₂via a Co(iii)-hydrido intermediate.In contrast to an O-bound Co^II-ONOH intermediate, where N–O bond homolysis of the ON-OH moiety generates H₂O₂ (Scheme 2, reaction (IV)),^26b the N-bound [Co-(NOOH)]⁺ intermediate decomposes to form NO₂⁻ and a Co(iii)-H transient species, arising from β-hydrogen transfer from the NOO–H moiety to the cobalt-center (Scheme 2, reaction (II)).^30a,c,32 The Co(iii)-hydrido species may generate H₂ gas either (a) by its transformation to the Co(ii)-nitrito complex (2) and H₂ gas as observed in the case of Co^III-H intermediate chemistry^30a,c,e−g as proposed in the chemistry of the Co^I complex with H⁺ reduction³¹ and other metal-hydrido intermediates³² and also explained in O₂ formation in PN chemistry^17,33 or (b) by the reacting with another [Co-(NOOH)]⁺ intermediate (Scheme 2, reaction (III)).Furthermore, we have confirmed the H₂ formation in the NOM reaction of 1 with OH⁻ by headspace gas mass spectrometry (Fig. 3a). Also, carrying out the reaction of 1 with NaOD leads to the formation of the [Co-(NOOD)]⁺ intermediate, which then transforms to a Co^III-D transient species. Further, as described above, the Co^III-D species releases D₂ gas, detected by headspace gas mass spectrometry (Fig. 3b), which evidently established that H₂ gas formed in the reaction of 1 with OH⁻. In this regard, we have proposed that in the first step of this reaction, the nucleophilic addition of OH⁻ to {Co-NO}⁸ generates a transient N-bound [Co-(NOOH)]⁺ intermediate that is generated by an internal electron transfer to Co^III (Scheme 2, reaction (I)). By following the mechanism proposed in the case of Co^III-H,^30a−c O₂,¹⁵ and H₂O₂(ref. 26b) formation, we have proposed the sequences of the NOM reaction of 1, which leads to the generation of Co^II-nitrito and H₂ (Scheme 2, reaction (I)–(III) and Scheme 3). In the second step, O–H bond homolytic cleavage generates a Co^III-H transient species + NO₂⁻via a β-hydrogen elimination reaction of the [Co-(NOOH)]⁺ intermediate.³² The Co^III-H intermediate may undergo the following reactions to generate H₂ gas and Co^II-nitrito either (a) by the natural decomposition of the Co^III-H transient species to generate H₂,^30a,c,e−g or (b) by the H-atom abstraction from another [Co-(NOOH)]⁺ intermediate (Scheme 3). Also, to validate our assumption that the reaction goes through a plausible N-bound [Co-(NOOH)]⁺ intermediate followed by its transformation to the Co^III-H species (vide supra), we have performed the reaction of 1 with NaOH/NaOD (in 1 : 1 ratio). In this reaction, we have observed the formation of a mixture of H₂, D₂, and HD gases, which indicates clearly that the reaction goes through the formation of Co^III-H and Co^III-D transient species via the aforementioned mechanism (Fig. 3c). This is the only example where tracking of the H atoms has confirmed the H₂ generation from an N-bound NOO–H moiety as proposed for H₂ formation from Co^III-H.³⁰Open in a separate windowFig. 3Mass spectra of formation of (a) H₂ in the reaction of 1 (5.0 mM) with NaOH (5.0 mM), (b) D₂ in the reaction of 1 (5.0 mM) with NaOD (5.0 mM), (c) D₂, HD, and H₂ in the reaction of 1 (5.0 mM) with NaOD/NaOH (1 : 1), and (d) H₂ in the reaction of 1 (5.0 mM) with NaOH in the presence of 2,4 DTBP (50 mM).Open in a separate windowScheme 3NOM reaction of complex 1 in the presence of OH⁻, showing the different steps of the reaction.While, we do not have direct spectral evidence to support the formation of the transient N-bound [Co-(NOOH)]⁺ intermediate and its decomposition to the Co^III-H transient species via β-hydrogen transfer from the NOOH moiety to the cobalt center, support for its formation comes from our finding that the reactive hydrogen species can be trapped by using 2,4-di-tert-butyl-phenol (2,4-DTBP).³⁴ In this reaction, we observed the formation of 2,4-DTBP-dimer (2,4-DTBP-D, ∼67%) as a single product (ESI, ES, and Fig. S7^†). This result can readily be explained by the H-atom abstraction reaction of 2,4-DTBP either by [Co-(NOOH)]⁺ or Co^III-H, hence generating a phenoxyl-radical and 3 with H₂ (Fig. 3d and Scheme 2, reaction (a)). Also, we have detected H₂ gas formation in this reaction (ESI,^† ES, and Fig. 3d). In the next step, two phenoxyl radicals dimerized to give 2,4-DTBP-dimer (Scheme 2c, reaction (II)). Thus, the observation of 2,4-DTBP-dimer in good yield supports the proposed reaction mechanism (Scheme 2, reaction (a) and (b)). Further, the formation of 2,4 DTBP as a single product also rules out the formation of the hydroxyl radical as observed in the case of an O-bound nitrous acid intermediate.^26bFurthermore, we have explored the NOM reactivity of 1 with Na₂O/15-crown-5 (as the O²⁻ source) and observed the formation of the Co^II-nitrito complex (3) via a plausible Co^I-nitro (4) intermediate (Scheme 1, reaction (IIa); also see the ESI^† and ES); however, 1 was found to be inert towards H₂O (Scheme 1, reaction (III); also see the ESI, ES and Fig. S8^†). The product obtained in the reaction of 1 with O²⁻ was characterized by various spectroscopic measurements.^15,26b The UV-vis absorption band of 1 (λ_max = 370 nm) disappears upon the addition of 1 equiv. of Na₂O and a new band (λ_max = 535 nm) forms, which corresponds to 3 (ESI, Fig. S9^†). The FT-IR spectrum of the isolated product of the above reaction shows a characteristic peak for Co^II-bound nitrite at 1271 cm⁻¹, which shifts to 1245 cm⁻¹ when exchanged with ¹⁵N-labeled-NO (¹⁵N¹⁶O) (ESI, ES, and Fig. S10^†), clearly indicating the generation of nitrite from the NO ligand of complex 1.^26b The ESI-MS spectrum recorded for the isolated product (vide supra) shows a prominent ion peak at m/z 333.1, and its mass and isotope distribution pattern matches with [(12-TMC)Co^II(NO₂)]⁺ (calc. m/z 333.1) (ESI, Fig. S11^†). Also, we quantified the amount of 3 (85 ± 5%) by quantifying the amount of nitrite (85 ± 5%) using the Griess reagent test (ESI, ES, and Fig. S6^†).In summary, we have demonstrated the reaction of Co^III-nitrosyl, [(12-TMC)Co^III(NO⁻)]²⁺/{CoNO}⁸ (1), with mono-oxygen reactive species (O²⁻, OH⁻ and H₂O) (Scheme 1). For the first time, we have established the clear formation of a Co^II-nitrito complex, [(12TMC)Co^II(NO₂⁻)]⁺ (3), and H₂ in the reaction of 1 with one equivalent of OH⁻via a transient N-bound [Co-(NOOH)]⁺ (2) intermediate. This [Co-(NOOH)]⁺ intermediate undergoes the O–H bond homolytic cleavage and generates a Co^III-H transient species with NO₂⁻, via a β-hydrogen elimination reaction of the [Co-(NOOH)]⁺ intermediate, which upon decomposition produces H₂ gas. This is in contrast to our previous report, where acid-induced nitrite reduction of 3 generated 1 and H₂O₂via an O-bound Co^II-ONOH intermediate.^26b Complex 1 was found to be inert towards H₂O; however, we have observed the formation of 3 when reacted with O²⁻. It is important to note that H₂ formation involves a distinctive pathway of O–H bond homolytic cleavage in the [Co-(NOOH)]⁺ intermediate, followed by the generation of the proposed Co^III-H transient species (Co^II + 1/2H₂)³⁰ prior to H₂ evolution as described in Co^I chemistry with H⁺ in many different ligand frameworks.³¹ The present study is the first-ever report where the base induced NOM reaction of Co^III-nitrosyl (1) leads to Co^II-nitrito (3) with H₂ evolution via an N-bound [Co-(NOOH)]⁺ intermediate, in contrast to the chemistry of O-bound Co^II-ONOH^26b, hence adding an entirely new mechanistic insight of base induced H₂ gas evolution and an additional pathway for NOM reactions.     

Enantioselective hydrogenation of annulated arenes: controlled formation of multiple stereocenters in adjacent rings

We report a method for the enantioselective hydrogenation of annulated arenes using 4H-pyrido[1,2-a]pyrimidinones as substrates. The method selectively generates multiple stereocenters in adjacent rings leading to architecturally complex motifs, which resemble bioactive molecules. The mechanistic study of the stereochemical outcome revealed that the catalyst is able to overcome substrate stereocontrol providing all-cis-substituted products predominantly. In a sequential protocol, a matching interaction between catalyst and substrate stereocontrol is achieved that facilitates diastereo- and enantioselective access to trans-products.

We report a method for the enantioselective hydrogenation of annulated arenes using 4H-pyrido[1,2-a]pyrimidinones as substrates.

Chiral, saturated carbo- and heterocycles are important structural elements of secondary metabolites and active ingredients in drugs and agrochemicals.¹ Compared to classical synthetic routes that utilize prefunctionalized precursors,² enantioselective arene hydrogenation provides a more direct route to such motifs from readily accessible (hetero-) aromatic substrates.³ With current hydrogenation technologies, a variety of monocyclic arenes such as pyridines,⁴ furans,⁵ thiophenes,⁶ and annulated arenes such as quinolines and isoquinolines,⁷ indoles,⁸ and naphthalenes⁹ can be hydrogenated with high enantioselectivity. However, in all previous reports on the enantioselective hydrogenation of annulated arenes (1) the resulting products (2) contain only a single saturated ring with at least one aromatic sextet being preserved in the product, and consequently, with stereocenter(s) only being formed in one ring (Fig. 1A).¹⁰ An enantioselective hydrogenation of multiple annulated aromatic rings would enable the generation of multiple stereocenters at adjacent rings in a single step, leading to desirable, architecturally complex and natural product-like motifs. Herein, we report the first such example using N-bridged 4H-pyrido[1,2-a]pyrimidinone (pyrido-pyrimidinone) substrates (3, Fig. 1B).^10a–d Pyrido-pyrimidinones, and their saturated analogs, are featured in bioactive molecules (Fig. 1C). For instance, Roche''s Risdiplam (5), an inhibitor against spinal muscular atrophy, includes a pyrido-pyrimidinone unit¹¹ and a semi-reduced tetrahydropyrido-pyrimidinone is a part of risperidone (6), a blockbuster drug against schizophrenia, while octahydropyrido-pyrimidinones (4) are structurally closely related to quinolizidine alkaloids such as lupanine (7). To date, only a small number of octahydropyrido-pyrimidinones (4) have been accessed with the majority resulting from the hetero-geneous, racemic hydrogenation of tetrahydropyrido-pyrimidinones (8). To the best of our knowledge, enantioenriched octahydropyrido-pyrimidinones have never been accessed by enantioselective catalysis.¹²Open in a separate windowFig. 1(A) Previously: only one ring of annulated arenes reduced by enantioselective hydrogenation. (B) This work: enantioselective hydrogenation of both rings of bicyclic aromatic pyrido-pyrimidinones (3). (C) Biologically active examples/analogues of pyrido-pyrimidinones and their hydrogenated products.We realized that an enantioselective hydrogenation of the readily accessible pyrido-pyrimidinones would in principle offer the most direct approach towards these valuable product motifs, provided that the chemoselectivity could be controlled and potential side products, such as tetrahydro- (8), hexahydropyrido-pyrimidinones 9 and 10, and the presumably unstable hemiaminal 11 resulting from amide reduction, could be avoided (Fig. 2).¹³ Furthermore, we rationalized that up to four different diastereomers could result from the creation of three independent stereocenters, making the simultaneous control of chemo-, diastereo-, and enantioselectivity a daunting challenge. Nevertheless, we were ideally positioned to overcome this challenge given the high reactivity, enantioselectivity, and broad tolerance of functional groups shown by the chiral ruthenium-bisNHC catalyst system developed by our group (12).^14,15 Indeed, we observed that catalytic hydrogenation of model substrate 2,7-dimethylpyrido-pyrimidinone (3a) proceeded with near-complete chemoselectivity under a variety of reaction conditions (Table S1^†). Under the optimized conditions, the desired octahydropyrido-pyrimidinone was obtained in 92% yield as a mixture of just two from the possible four diastereomers (63 : 37 diastereomeric ratio (dr)) and with excellent 96% enantiomeric excess (ee) for the major diastereomer (Table S1,^† entry 4). The evaluation of other substitution patterns of the pyrido- pyrimidinone using the respective dimethyl-substituted substrates revealed that the 2,8-disubstituted product can also be obtained with good chemo-, enhanced diastereo-, and moderate enantioselectivity. However, substrates with 2,6- and 2,9-substitution patterns were less reactive. Only the eastern pyridine ring was reduced in 48% and 82% yield, respectively (Table S2^†). As such we decided to focus on substrates with a 2,7-substitution pattern when exploring the scope of this reaction (Fig. 3). Next, the scope was evaluated (Fig. 3). Substrates without a substituent in the C2-position (3b), and substrates containing sterically more hindering substituents such as a benzyl or an iso-propyl group (3c,d) in the C2-position all delivered the products in high yields as a mixture of just two diastereomers with moderate dr and excellent ee values for the major diastereomer. The dr values increase with the steric size of the C2-substituent. A phenyl substituent in the C2-position rendered the (eastern) pyrimidinone ring inactive, thus chemoselectively providing tetrahydropyrido-pyrimidinone 8e with moderate ee under the standard reaction conditions including the reaction time of 48 h. Contrastingly, the introduction of aryl groups in the 7-position of the (western) pyridine ring was well-tolerated. Consequently, octahydropyrido-pyrimidinones carrying aryl groups such as phenyl (4f), sterically demanding ortho-tolyl (4g), and other aryl groups bearing various synthetically useful functional groups such as NHBoc, OCF₃, or Cl in the para-position (4h–j) can all be accessed in high yields and ee values. Curiously, while 7-para-fluorophenyl-substituted product 4k was obtained in high yield, the enantioselectivity was reduced in comparison to its chlorine-containing counterpart 4j. Products without a substituent in the 7-position (4l–n) were isolated in good yields as single diastereomers with good to high ee.Open in a separate windowFig. 2Chemo- and stereoselectivity challenges.Open in a separate windowFig. 3Scope. Combined isolated yields of both diastereomers are reported. The dr was determined by NMR-spectroscopy. The major diastereomer (and enantiomer) was assigned by X-ray crystallographic analysis of both diastereomers of 4a. The assignment for the other products was conducted by analogy. The ee of the minor diastereomer was measured in all cases (see ESI^†). ^aTraces of an unknown impurity could not be separated from the product. ^bProtection with trifluoroacetic anhydride. ^c80 bar H₂ used. SINpEt: 1,3-bis(1-(naphthalene-1-yl)ethyl)-4,5-dihydroimidazolylidene; TFA: trifluoroacetate.Intrigued by those observations, we proceeded to study the stereochemical mechanism of the reaction. First, we determined the reaction progress of the hydrogenation of model substrate 3a over time (Fig. 4A). After a reaction time of 1.5 h, tetrahydropyrido-pyrimidinone 8a is almost the exclusive species before its relative abundance decreases with a coinciding increase of the proportion of octahydropyrido-pyrimidinone 4a. Notably, 7-methyl- and 7-phenyl-substituted intermediates 8a and 8f were isolated in good yields and with moderate enantiomeric excess (8a: 84% yield, 63% ee; 8f: 50% yield, 62% ee). The reaction is complete after 8 h, leaving octahydrogenated product 4a as the exclusive species. Hence 8a must be an intermediate in the hydrogenation of 3a to 4a. No further signals, other than those allocated to the substrate 3a and tetrahydrointermediate 8a, were observed in the enone or α-ketone regions at any of the time points. Furthermore, the C2 and C9a C–H bonds are cis-aligned in both diastereomers of the 2,7-dimethylproduct 4a (confirmed by X-ray crystallography, see ESI^†). A cis alignment would be enforced by a non-interrupted coordination of the catalyst during the hydrogenation of the (eastern) pyrimidinone ring. The codependence of the configurations of C2 and C9a would result in the formation of only two diastereomers in the hydrogenation of 7-substituted substrates and only a single diastereomer in the hydrogenation of 7-unsubstituted substrates, which is in agreement with the observations made during the exploration of the scope (Fig. 4B).¹⁶ Both observations suggest that enone 9a and imine 10a are not intermediates in this reaction. Finally, we determined that two separate effects cause the (relatively low) diastereomeric ratios of 7-substituted octahydropyrido-pyrimidinones. Firstly, the enantiocontrol over the initially formed C7-stereocenter is only moderate (8a: 63% ee after 1.5 h, Fig. 4B). Secondly, the ee of the intermediates 8 increases with the reaction progress of further hydrogenation towards the octahydroproducts 4 (Fig. 4C). It follows that the minor enantiomer of chiral intermediate 8 is converted into octahydro-product 4 more rapidly than the major enantiomer, indicating a mismatched interaction between substrate and catalyst control. To further test this conclusion, isolated intermediate 8a was submitted to the enantioselective hydrogenation using the opposite enantiomer of the chiral catalyst in the second hydrogenation step (Fig. 4D).¹⁶ In this case, hydrogenation proceeds with a matched interaction between catalyst and substrate control and provides the opposite diastereomer trans–cis-4a with excellent diastereo- and enantiocontrol. Notably even racemic arene hydrogenation rarely produces trans-products.¹⁸ The dr of all–cis-4a derived from the two-step hydrogenation (Fig. 4D, red equation) is higher than that of the standard one-pot protocol (75 : 25 vs 56 : 44 dr). This is a result of the increase of the ee of intermediate 8a as a function of the reaction time (the minor enantiomer is consumed faster to form 4a). The effective ee of 8a in the two-step protocol (isolated after 2 h) was higher than that of 8a in the standard protocol resulting in a higher dr of 4a.Open in a separate windowFig. 4Mechanistic studies. (A) Excerpt of the ¹H NMR spectrum of the reaction mixture as a function of the reaction time. (B) Key observations that suggest a two-step hydrogenation process. (C) ee of 8a as a function of the reaction time. (D) Sequential enantioselective hydrogenation of 3a using opposite enantiomers of the chiral catalyst.¹⁷     

Correction: Crystal structure and metallization mechanism of the π-radical metal TED

Radical electrons tend to localize on individual molecules, resulting in an insulating (Mott–Hubbard) bandgap in the solid state. Herein, we report the crystal structure and intrinsic electronic properties of the first single crystal of a π-radical metal, tetrathiafulvalene-extended dicarboxylate (TED). The electrical conductivity is up to 30 000 S cm⁻¹ at 2 K and 2300 S cm⁻¹ at room temperature. Temperature dependence of resistivity obeys a T³ power-law above T > 100 K, indicating a new type of metal. X-ray crystallographic analysis clarifies the planar TED molecule, with a symmetric intramolecular hydrogen bond, is stacked along longitudinal (the a-axis) and transverse (the b-axis) directions. The π-orbitals are distributed to avoid strong local interactions. First-principles electronic calculations reveal the origin of the metallization giving rise to a wide bandwidth exceeding 1 eV near the Fermi level. TED demonstrates the effect of two-dimensional stacking of π-orbitals on electron delocalization, where a high carrier mobility of 31.6 cm² V⁻¹ s⁻¹ (113 K) is achieved.

The molecular arrangement that enables metallic conduction in a single-component pure organic crystal is revealed by single-crystal X-ray diffraction.

Organic molecular solids are typically insulating due to their paired electrons in spatially localized s- and p-orbitals. The concept of charge-transfer (CT) between donor and acceptor¹ enabled the development of conducting molecular complexes (salts) including semiconducting perylene-bromine,² metallic tetrathiafulvalene (TTF)-tetracyano-p-quinodimethane (TCNQ),³ and polyacethylene doped with halogen molecules.⁴ A different strategy was proposed in the 1970s based on organic radicals with an open-shell electronic structure.⁵ π-Radicals such as neutral-,⁶ fully-conjugated⁷ and zwitterionic (betainic)⁸ molecules, with an unpaired electron in their singly occupied molecular orbital (SOMO), offered potential candidates. However, all these π-radicals were insulators or semiconductors with a finite bandgap, which is due to the SOMO being localized on an individual molecule. In the Mott–Hubbard model,⁹ the case of the π-radical solids can be described by the on-site Coulomb repulsion U being larger than the electronic bandwidth W (U/W > 1), in contrast to U/W < 1 in molecular metals like CT metal systems (Fig. 1a).Open in a separate windowFig. 1(a) Schematic representation of the electronic band structure of Mott–Hubbard insulators (left) and possible molecular metals (right), respectively. Solids formed from typical π-radicals possess large on-site Coulomb repulsion U compared with the electronic bandwidth W, resulting a finite bandgap (left). This requires a new mechanism to expand W and overcome U to achieve a metallic state at ambient pressure in π-radical crystals. (b) Molecular structure of the zwitterionic radical, tetrathiafulvalene-extended dicarboxylate (TED) with a symmetric intramolecular hydrogen bond.A straightforward approach to realize high conductivity in π-radical systems is to enhance the intermolecular interaction by applying high pressure.¹⁰ Bisdithiazolyl radical crystal achieved W ∼ 1 eV near the Fermi level and the room temperature conductivity σ_RT = 2 S cm⁻¹ under 11 GPa pressure.^10c An alternative route is to decrease the interatomic spacing by incorporating a metal ion. Introduction of a semimetal Se and intermolecular hydrogen bonding in a donor-type radical succeeded to improve a conductivity to σ_RT = 19 S cm⁻¹ but still required high pressure over 1 GPa for breaking its insulating character.^10d An organometallic compound with a transition metal, [Ni(tmdt)₂], by contrast, is known to form a three-dimensional (3D) Fermi surface with W = 0.48 eV and metallic conduction with σ_RT = 400 S cm⁻¹ at ambient pressure.¹¹ A breakthrough concept for expanding W at ambient pressure is desired for achieving metallization in pure organic π-radicals.Tetrathiafulvalene-extended dicarboxylate (TED) is an organic air-stable zwitterionic radical (Fig. 1b),¹² which was designed based on carrier generation induced by a stably-introduced protonic defect (–H⁺) in hydrogen-bonding molecules without adopting CT between multiple molecules.¹³ A polycrystalline film of TED exhibited metallic conduction at ambient pressure, but the mechanism has not been clarified yet due to lack of single crystal information.¹² Herein, we report the first crystal structure and intrinsic electronic properties of the recently grown single crystal TED. Structural analysis and quantum chemical simulations based on the single crystal reveal the origin of its metallic behavior.     

Crystals springing into action: metal–organic framework CUK-1 as a pressure-driven molecular spring

Mercury porosimetry and in situ high pressure single crystal X-ray diffraction revealed the wine-rack CUK-1 MOF as a unique crystalline material capable of a fully reversible mechanical pressure-triggered structural contraction. The near-absence of hysteresis upon cycling exhibited by this robust MOF, akin to an ideal molecular spring, is associated with a constant work energy storage capacity of 40 J g⁻¹. Molecular simulations were further deployed to uncover the free-energy landscape behind this unprecedented pressure-responsive phenomenon in the area of compliant hybrid porous materials. This discovery is of utmost importance from the perspective of instant energy storage and delivery.

Mercury porosimetry and in situ high pressure single crystal X-ray diffraction revealed the wine-rack CUK-1 MOF as a unique crystalline material capable of a fully reversible mechanical pressure-triggered structural contraction.

Reducing the world''s fossil fuel dependence is the focus of many global initiatives,¹ aiming to mitigate the effects of climate change through tapping into sustainable energy resources such as solar and wind power. However, increasing reliance on these renewable energy sources has introduced difficulties due to the offset between power availability and demand peaks. Complementary technologies are necessary to alleviate intermittent supply, such as peaking power plants, demand-side energy management, or large scale energy storage.² The latter is particularly desirable as it can decouple electricity production and consumption, however the lack of a “one size fits all” approach has led the scientific community to envisage unconventional energy storage strategies.One such avenue emerging in recent years is the storage of mechanical energy via the compression of a suitable stimuli-responsive system, either through the intrusion of a non-wetting fluid into hydrophobic porous frameworks,³ or by means of application of an external pressure on flexible materials.⁴The former approach, first pioneered using water intrusion in zeolites and silicas,¹¹ has recently been extended to small pore zeolitic imidazolate frameworks.¹² Unfortunately, besides requiring highly hydrophobic systems, water intrusion achieves a relatively low stored energy density,³ of around 3–25 J g⁻¹. The second strategy takes advantage of the compliant nature of bulk materials. Energy is stored through structural deformations, manifesting as continuous or sudden volume changes under external pressure. The energy stored in flexible materials over a compression/decompression cycle can be an order of magnitude higher compared to the values achieved using fluid intrusion in rigid porous systems.¹³ In theory, three types of pressure-induced structural behaviour can be envisioned for such a responsive system. If the structure contraction is non-reversible, all energy is dissipated and the system is categorized as a nano-shock absorber (Fig. 1b). For structural changes that are reversible upon decompression two families of system can be distinguished, i.e. a nano-damper (Fig. 1c) or an ideal nano-spring (Fig. 1d) when the p–V curves show hysteresis or fully overlap, respectively.¹⁴Open in a separate windowFig. 1(A) Schema of mechanical energy storage in compliant crystalline materials, implying a unit cell volume change between open (op) and contracted (cp) structures, and prototypical pressure-volume curves of stimuli-responsive materials under mechanical pressure for (B) nano-shock absorbers, exemplified by MIL-53(Al),⁵ MIL-53(Ga)-FA⁶ and ZIF-4(Zn),⁷ (C) nano-dampers e.g. MIL-53(Cr),⁸ MIL-47(V)⁹ and MIL-53(Al)-FA¹⁰ and (D) nano-springs, insofar exhibited exclusively by CUK-1 presented herein.Metal–organic frameworks (MOFs), a class of porous, crystalline materials comprised of metal vertices interconnected by organic linkers, are known to exhibit responsiveness to a variety of stimuli,^15,16 including external pressure.¹⁷ Recently, several frameworks of this family of hybrid materials have been shown to act as energy storing nano-dampers or energy dissipative nano-shock absorbers, as is the case for the highly flexible MIL-53(M)^5,8,10 and MIL-47(V)⁹ series and more recently ZIF-4(Zn)⁷ (see Fig. 1b and c for their related structural behaviours). In such flexible crystalline materials compression is associated with a displacive phase transition between distinct structures of differing unit cell volumes, denoted as open (op) and contracted (cp) forms^15,16 and illustrated in Fig. 1a, occurring reversibly or irreversibly for a nano-damper or nano-shock absorber, respectively. The considerable stored energy associated with this transition, in the range of 30–200 J g⁻¹ (up to 4 kJ g⁻¹ for shock absorbers¹⁸) is highly attractive from the perspective of mechanical energy storage. However, the hysteretic compression/decompression curve characterising known nano-damper MOFs leads to a partial loss of work energy, lowering the potential storage efficiency, as well as creating issues through heat dissipation. Insofar, the search for a ideal spring-like crystalline material, capable of reversible pressure-induced structural switching without any hysteresis (Fig. 1d) has been fruitless, precluding their applicability for efficient, high density energy storage applications. Herein, a subtle combination of Hg-porosimetry, high-pressure single crystal X-ray diffraction (SC-XRD) and molecular simulations reveals the 1D-channel CUK-1 (M, M = Co, Mg)¹⁹ MOF as the first compliant hybrid porous material with a spring-back mechanical breathing behaviour.Such unique mechanically-triggered structural response implies a continuous pore contraction/expansion between op and cp forms in a narrow pressure range of 280–290 MPa, accompanied by a unit cell volume change of 20.9%. This optimal scenario paves the way towards fast energy storage/delivery system of about 40 J g⁻¹. The channel-like CUK-1(M) composed of chains of μ₃-OH/O edge and vertex sharing metal octahedra (M = Co,¹⁹ Mg²⁰) coordinated by bidentate 2,4-pyridinedicarboxylic ligands, recently emerged as an attractive porous material owing to its promising sorption performance combined with environmentally-friendly hydrothermal synthesis and high thermal and chemical stability.^20–22 Its wine-rack topology and its relatively rigid behaviour upon guest adsorption are reminiscent to that of MIL-47(V) a MOF which interestingly underwent a hysteretic, reversible structural contraction upon exerting an external pressure of 125 MPa,⁹ associated with a stored/delivered energy of 33 J g⁻¹. Inspired by our previous findings on MIL-47(V), we deliberately explored the pressure-induced structural behaviour of CUK-1 in its isostructural Co and Mg forms. MOF synthesis was performed according to the protocol detailed in ESI.^† Phase purity was confirmed by powder XRD (Fig. S3, S4 and Table S1^†) while their textural features, including BET area and pore volume, were found to match previously reported data.^19,20Mercury intrusion curves were recorded on the powder samples up to a maximum of 413 MPa as shown for CUK-1(Co) in Fig. 2, its Mg variant being reported in Fig. S6, ESI,^† together with full experimental details. A substantial amount of Hg intrudes at low pressure (<10 MPa), due to compaction of the crystals and filling of inter-particle porosity. This is followed by a sudden volume change at 281 MPa where a sharp step is observed (see inset of Fig. 2). By analogy with the conclusions previously drawn for the series of MIL-53(M)/MIL-47(V) frameworks,^5,8–10 this intruded Hg volume increase is associated with a structural contraction of CUK-1(Co), as its channel size (approx. 6.6 Å) is an order of magnitude below the pore dimension where non-wetting mercury can intrude in this pressure range (at 52 Å). The extrusion curve shows a near-perfect overlap, indicating that the framework behaves as an ideal spring, with no hysteresis between the intrusion/extrusion branches.Open in a separate windowFig. 2Sequential mercury intrusion–extrusion curves on CUK-1(Co) powder, in blue line and red, respectively. Line is a guide for eye. Volume below 1 MPa corresponds to powder compaction and intercrystallite void filling. Dotted horizontal lines demarcate contraction lower and upper bounds. Inset highlights the intrusion step in a linear scale with the op/cp contraction marked with an arrow.Moreover, this behaviour is highly repeatable, as confirmed by four consecutive pressure cycles (in Fig. S5^†). Interestingly, the same behaviour also holds true for CUK-1(Mg) (Fig. S6^†), with a similar intrusion pressure of 288 MPa. Since the two metal ions show relatively similar ionic radius (Co²⁺: 1.50(7) Å and Mg²⁺: 1.41(7) Å),²⁴ the averaged metal–oxygen distance is nearly identical in their corresponding coordination: spheres: (Co–O: 2.107(20) Å and Mg–O: 2.073(20) Å). Such analogous metal-linker bond strength is most likely at the origin of the very similar pressure-induced response of the two materials. The high transition pressure of CUK-1(Co) underpins the inability of guest adsorption to induce a breathing effect as observed previously.^20,21 Indeed, the adsorption stresses encountered throughout guest insertion are simply insufficient to overcome the energetic penalty of transition.^14,25 The 0.143 mL g⁻¹ volume change associated with the observed step in the CUK-1(Co) intrusion curve corresponds to a 20.9% change in unit cell volume, lower than in the similar phenyl-based MIL-47(V) of 43%.⁹ However, the stored energy calculated through W = P × ΔV is 40 J g⁻¹, 20% larger than the value reported for MIL-47(V)⁹ of 33 J g⁻¹. Here, the higher pressure of CUK-1(Co) switching, 281 MPa vs. 125 MPa for MIL-47(V) balances out the ΔV term. Moreover, owing to its relatively dense framework, the volumetric energy density of CUK-1(Co) remains attractive when compared to water intrusion systems (Table S4^†).Considering an initial unit cell volume for the CUK-1(Co) op form of 2467 ų from PXRD (see ESI^†), the resulting cp form is estimated to exhibit a unit cell volume of 1950 ų, based on the Hg intruded volume increase at 281 MPa. In order to directly observe the contracted form and identify the mechanism underpinning these intriguing dynamics, high pressure SC-XRD experiments were carried out in a membrane diamond anvil cell (mDAC). Individual CUK-1(Co) crystals were placed in a gasket between the polished diamonds of the mDAC, and immersed in a hydrostatic pressure transmitting medium of silicone oil AP-100, with a gold flake used to monitor inner mDAC pressure (full single crystal synthesis conditions and SC-XRD methodology available in the ESI^†).At ambient pressure, the indexed unit cell volume of the initial op form of CUK-1(Co) is nearly identical (2492 ų) to that of the previously reported²⁰ dehydrated monoclinic phase (2466.72 ų). Upon increasing DAC pressure to around 0.3 GPa, a volume contraction to the cp phase begins, which is in line with Hg porosimetry experiments. Reflections obtained from integrated 2D diffraction images were used to solve the pressure-induced structure through a dual space recycling algorithm in an expanded P1 setting, then further refined on F² using the SHELX suite²⁶ (complete data treatment methodology available in the ESI^†). The structure maintains the same C2/c space group throughout the transition between the two forms, and as such the spring-like dynamics of the framework can be described as a continuous contraction in a narrow pressure range. Above 0.5 GPa, the cp form is attained, with further pressure application leading to a linear decrease of its unit cell volume by 4% up to 1.8 GPa (Fig. 3).Open in a separate windowFig. 3Evolution of the CUK-1(Co) unit cell volume determined through indexation of Bragg reflections as a function of applied pressure as recorded in a DAC. Unit cell parameters corresponding to each pressure point can be found in Table S5, ESI.^†The unit cell dimensions of the solved cp form at 0.5 GPa are provided in Table 1, alongside as-indexed pristine op form parameters with Fig. 4a illustrating the two structures. The anisotropic transition is similar in nature to that of MIL-53(M)/MIL-47(V), characterised by a compaction in the b-direction (from approx. 13 Å to 9 Å) and an elongation along the a-axis (from 18 to nearly 20 Å). The change in the c-parameter is minimal, with only a slight increase, as it lies in the plane of the highly rigid octahedrally coordinated metal chains. A lowering of the angle (from 103 to 99°) is also observed, as the 1D parallel pores are straightened via the linker-induced torsion. A table comparing specific atomic distances, angles and torsions in the two forms is available in Table S7, ESI.^† The unit cell volume of the identified cp phase at 0.5 GPa of 1972 ų is only slightly higher than the value estimated from porosimetry measurements (as 1950 ų). We attribute this offset to the different interactions of the crystal surface with the respective pressure transmitting medium (mercury vs. silicone oil), as observed previously.⁵Crystallographic data of the pristine (op) and high pressure (cp) phases as determined from the CUK-1(Co) SC-XRD

Chiral information harvesting in helical poly(acetylene) derivatives using oligo(p-phenyleneethynylene)s as spacers

A chiral harvesting transmission mechanism is described in poly(acetylene)s bearing oligo(p-phenyleneethynylene)s (OPEs) used as rigid achiral spacers and derivatized with chiral pendant groups. The chiral moieties induce a positive or negative tilting degree in the stacking of OPE units along the polymer structure, which is further harvested by the polyene backbone adopting either a P or M helix.

A chiral harvesting transmission mechanism is described in poly(acetylene)s bearing oligo(p-phenyleneethynylene)s (OPEs) used as rigid achiral spacers and derivatized with chiral pendant groups.

During the last years, dynamic helical polymers have attracted the attention of the scientific community due to the possibility of tuning the helical sense and/or the elongation of the helical structure by using external stimuli.^1–14In the case of a chiral dynamic helical polymer, modifications in its structure—helical sense enhancement or helix inversion—arise from conformational changes induced at its chiral pendants—usually, with just one stereocenter—, by stimuli such as variations in solvent polarity or temperature, the addition of certain ions, and so on (Fig. 1a).¹⁵ On the other hand, if a helical polymer is achiral (i.e., bearing achiral pendants), the chiral amplification phenomena can emerge from interactions between the polymer and external chiral molecules.¹⁶ In both the above cases, the changes produced in the helical structures are related to the spatial dispositions adopted by the substituents or associated species at the pendant groups.^17–19Open in a separate windowFig. 1Several scenarios depicting conceptual representations of the transmission of chiral information. (a) Helical switch via chiral tele-induction. (b) Effect of distance on chiral tele-induction from multichiral pendants. (c) Helicity controlled by the conformational composition of achiral spacers.A step forward in the helical sense control of poly(phenylacetylene)s (PPA)s is to study different mechanisms of transmission of chiral information from the pendant to the polyene backbone by introducing achiral spacers. The goal is to demonstrate how far it is possible to place the chiral center and still have an effective chiral induction on the polyene backbone. Therefore, transmission of the chiral information from a remote position can occur through space, thus overpassing the distance generated by the spacer—tele-induction—(Fig. 1b),^20–28 or through the achiral spacer itself, producing in it a preferred structure, such as a helical structure and where the orientation of the achiral helix is further transmitted to the polyene backbone—conformational switch—(Fig. 1c).^29–31For the first mechanism—chiral tele-induction—, both flexible and rigid spacers have been designed.^20–28 In all cases, supramolecular interactions, such as H bonding or π–π stacking, generate organized structures. As a result, the chiral center is located into a specific orientation, producing an effective helical induction. Additionally, those studies allow evaluating how distances and sizes have an effect on this phenomenon.In the second strategy, the helix induction is transmitted through conformational changes along an achiral spacer which is harvested by the polyene. For instance, an achiral peptide or an achiral polymeric helix derivatized at one end with a chiral residue and linked to the polymer main chain at the other end. In such cases, changes in the absolute configuration or even just a conformational change at the chiral center can induce an opposite helical structure into the achiral spacer, which in turn will be harvested by the polymer main chain (Fig. 1c).^29–31Herein we will demonstrate another remote chiral induction mechanism based on a different chiral harvesting process. In this case, the chiral center does not produce a conformational change at the achiral spacer, but affects its array within the helical scaffold. Thus, to perform these studies we decided to introduce the use of oligo(p-phenyleneethynylene)s (m = 1, 2, 3) (OPEs) as rigid spacers to separate the distant chiral center from the polyene backbone. These OPE units have been used in the formation of benzene-1,3,5-tricarboxamide (BTA) based supramolecular helical polymers, demonstrating their ability to stack with a certain tilting degree commanded by the chiral center.^32–34Hence, in our design, the chiral moiety will determine the supramolecular chiral orientation of the OPE groups used as spacers, which is further harvested by the polyene backbone. The overall process yields a helix with a preferred screw sense (Fig. 2).Open in a separate windowFig. 2Conceptual side view and top view of the chiral information transmission mechanism from stereocenters at the far end of oligo(p-phenyleneethynylene) spacers to the polyene backbone via chiral harvesting.To perform these studies, we used as model compounds two PPAs—poly-(R)-1 and poly-(S)-1—derived from the 4-ethynylanilide of (S)- and (R)-α-methoxy-α-phenylacetic acid (MPA, m-(S/R)-1), whose helical structures and dynamic behaviors have been deeply studied by our group—poly-(R)-1 and poly-(S)-1—(Fig. 3).^35–46 By using these polymers as reference materials, four novel PPAs were designed introducing two OPE spacers—4-[(p-phenyleneethynylene)_n]ethynylanilide (n = 1, 2)—between the phenyl acetylene group and the (S)- or (R)-α-methoxy-α-phenylacetic acid (MPA) chiral group. Thus, monomers m-(S)- and m-(R)-2 and m-(S)- and m-(R)-3 (Fig. 3a) were prepared and submitted to polymerization by using a Rh(i) catalyst poly-(S)- and poly-(R)-2 and poly-(S)- and poly-(R)-3 (Fig. 3b) were obtained in high yield and showed Raman spectra characteristic of cis polyene backbones (see Fig. S11 and S12^†).Open in a separate windowFig. 3(a) Monomers and (b) polymers synthetized in this study.X-ray structures of the monomers show a preferred antiperiplanar (ap) orientation between the carbonyl and methoxy groups (O C–C–OMe) for m-(R)-2 and m-(S)-3, whereas in the case of m-(S)-1 a synperiplanar (sp) geometry is favoured (Fig. 4a).³⁵ In complementary studies, CD spectra of monomers m-(S)-[1–3] in CHCl₃ show negative Cotton effects, indicative of major ap conformations in solution (Fig. 4b),³⁵ further corroborated by theoretical calculations (see Fig. S10^†). Interestingly, the maximums of the Cotton effects in CD undergo a bathochromic shift—from 266 nm in m-1 to 327 nm in m-3—due to a larger conjugation of the π electrons (from the anilide to the alkyne group) when the length of the spacer increases (Fig. 4b).Open in a separate windowFig. 4(a) X-ray structures of m-(S)-1, m-(R)-2 and m-(S)-3. (b) CD traces of m-(S)- and m-(R)-1; m-(S)- and m-(R)-2; m-(S)- and m-(R)-3 in CHCl₃ (0.1 mg mL⁻¹). (c) CD spectra for poly-(S)- and poly-(R)-1 in CHCl₃ (0.1 mg mL⁻¹); poly-(S)- and poly-(R)-2 in DMSO (0.1 mg mL⁻¹); poly-(S)- and poly-(R)-3 in DMSO (0.1 mg mL⁻¹).CD studies of the polymer series bearing OPE spacers—poly-(R)- and poly-(S)-[2–3]—in different solvents show the formation of a PPA helical structure with a preferred helical sense, while the parent polymer, poly-1, devoid of the OPE unit, has a poor CD. This is a very interesting phenomena that indicates that the OPE spacers work as transmitters of the chiral information from remote chiral centers to the polyene backbone—placed at 1.7 nm for poly-2 and at 2.4 nm for poly-3—(Fig. 4a). These large distances between the chiral center and the polymer main chain mean that other mechanisms of chiral induction, such as chiral tele-induction effect, should be almost null in these cases.In these two polymers (poly-2 and poly-3), the chiral information transmission mechanism must occur in different sequential steps. First, the chiral centers possessing a major (ap) conformation induce a certain tilting degree (θ) in the achiral spacer array. This step resembles the helical induction mechanism found in supramolecular helical polymers bearing OPE units.^32–34 Next, the chiral array induced in the OPE units is harvested by the polyene backbone, resulting in an effective P or M helix induction (Fig. 2).^34,47Additional structural studies were carried out in poly-(S)-2 and poly-(S)-3 to obtain an approximated secondary structure of these polymers and determine their dynamic behaviour.From literature it is known that the conformational equilibrium of poly-1 can be altered in solution by the presence of metal ions. The addition of monovalent ions (e.g., Li⁺) stabilizes the ap conformer at the pendant group by cation–π interactions, while divalent ions (e.g., Ca²⁺) stabilize the sp conformations by chelation with the methoxy and carbonyl groups.^36,38,39,43 As a result, both the P or M helical senses can be selectively induced in poly-1 by the action of metal ions.Therefore, we decided to add different perchlorates of monovalent and divalent metal ions to solutions of poly-(S)-2 and poly-(S)-3 with the aim of determining the conformational composition at the pendant groups. Thus, when monovalent metal ions (Li⁺, Ag⁺ and Na⁺) are added to a chloroform solution of poly-(S)-2, a chiral enhancement is observed (Fig. 5d for Li⁺ and Fig. S16^† for Na⁺ and Ag⁺). IR and ⁷Li-NMR studies show that those ions stabilize the ap conformer at the pendant group in a similar fashion to poly-1, this is by coordination to the carbonyl group of the MPA (Fig. 5g) and the presence of a cation–π interaction with the aryl ring of the chiral (|Δδ| ⁷Li ca., 3.75 ppm) (Fig. 5f and ESI^†). Therefore, addition of Li⁺ produces a larger number of pendant groups with ap conformation among poly-2, which triggers a chiral enhancement effect through a cooperative process.Open in a separate windowFig. 5(a) Conceptual representation of the chiral information harvesting and top view of the 3D model for poly-(S)-2. (b) CD spectra of poly-(S)-2 (0.2 mg mL⁻¹) in DMSO vs. calculated ECD spectra. Full width at half-maximum (FWHM) equals 20 nm. (c) Low-resolution AFM image from a poly-(S)-2 monolayer and profile depicting the chain separation of the yellow highlighted area in the AFM image. (d) CD spectra showing the chiral enhancement after the addition of Li⁺ (50 mg mL⁻¹, THF) to a poly-(S)-2 solution (0.1 mg mL⁻¹, THF). (e) CD trace of poly-(S)-2 before and after the addition of a Ca²⁺ solution (50 mg mL⁻¹, THF). (f) ⁷Li-NMR spectra substantiating the cation–π interaction. (g) IR shifts observed for carbonyl and methoxy groups after the addition of LiClO₄ and Ca(ClO₄)₂ (50 mg mL⁻¹, THF) to a poly-(S)-2 solution (3 mg mL⁻¹, CHCl₃). The coordination modes of the MPA moiety with Li⁺ and Ca²⁺ are shown vertically in the middle of the figure.On the contrary, the addition of perchlorates of divalent metal ions, such as Ca²⁺and Zn²⁺, produced an inversion of the third Cotton band—310 nm—associated to the MPA moiety and the disappearance of both first and second Cotton effects (Fig. 5e for Ca²⁺ and Fig. S17^† for Zn²⁺). This is a very interesting outcome because, although the conformational equilibrium at the MPA group changes from ap to sp after the addition of Ca²⁺, the number of pendant groups with sp conformation do not reach the number needed to trigger the helix inversion process and in fact, a mixture of P and M helices at the polyene backbone is obtained.The helical structures adopted by both polymer systems, PPAs (poly-1) and poly[oligo(p-phenyleneethynylene)phenylacetylene]s (POPEPAs) (poly-2 and poly-3), are defined by two coaxial helices, one formed by the polyene backbone (internal helix, CD active) and the other constituted by the pendants (external helix, observed by AFM).These two helices can rotate in either the same or the opposite sense, depending on the dihedral angle between conjugated double bonds. Thus, internal and external helices rotate in the same direction in cis-cisoidal polymers, while they rotate in opposite directions in cis-transoidal ones.^14,42,48,49In order to find out an approximated helical structure for poly-(S)-2, DSC studies were performed. The thermogram shows a compressed cis-cisoidal polyene skeleton (see Fig. S13a^†), similar to the one obtained for poly-1.⁴² Moreover, AFM studies on a 2D crystal of poly-(S)-2 did not produce high-resolution AFM images, although some parameters such as helical pitch (c.a., 2.8 nm) and packing distance between helices of (c.a., 6 nm) could be extracted from the well-ordered monolayer analyzed (Fig. 5c).Previous structural studies in PPAs found that it is possible to correlate the internal helical sense with the Cotton band associated to the polyene backbone—CD (+), P_int; CD (−), M_int—.^50,51 Herein, the positive Cotton effect observed for the polyene backbone [CD_{365 nm} = (+)] in poly-(S)-2 is indicative of a P orientation of the internal helix, which correlates with a P orientation of the external helix in a cis-cisoidal polyene scaffold. To summarize, DSC, AFM and CD studies agree that poly-(S)-2 is made up of a cis-cisoidal framework with P_int and P_ext helicities (Fig. 5a).Computational studies [TD-DFT(CAM-B3LYP)/3-21G] were carried out on a P helix of an n = 9 oligomer of poly-(S)-2, possessing a cis-cisoidal polyene skeleton (ω₁ = +50°, ω₃ = −40°) and an antiperiplanar orientation of the carbonyl and methoxy groups at the pendants. The theoretical ECD spectrum obtained from these studies (Fig. 5b and see ESI^† for additional information) is in good agreement with the experimental one, indicating that our model structure is a good approximation of the helical structure adopted by poly-(S)-2.Next, a similar set of DSC and AFM studies were carried out for poly-(S)-3, that bears an OPE spacer with n = 2. The data showed that this polymer presents a compressed cis-cisoidal polyene skeleton, similar to those obtained for poly-1 and poly-2 (see Fig. S13b^†), with a helical pitch of 3.8 nm and a P_ext helical sense (Fig. 6a and c).Open in a separate windowFig. 6(a) Conceptual representation of the chiral information harvesting and top view of the 3D model for poly-(S)-3. (b) CD spectrum of poly-(S)-3 in THF (0.2 mg mL⁻¹) and comparison to the calculated ECD spectra. Full width at half-maximum (FWHM) equals 20 nm. (c) AFM image obtained from a poly-(S)-3 monolayer. (d) CD traces for poly-(S)-3 in THF polymerized at different temperatures.UV studies indicate that, in poly-(S)-3, the polyene backbone absorbs at ca. 380 nm, coincident with the first Cotton effect, that is positive (see Fig. S15b^†). Therefore, it reveals that poly-(S)-3 adopts a P_int helicity (Fig. 6b). Thus, as expected for cis-cisoidal scaffolds, the orientations of the two coaxial helices are coincident.Computational studies [TD-DFT(CAM-B3LYP)/3-21G] were carried out on a P helix of an n = 9 oligomer of poly-(S)-3, possessing a cis-cisoidal polyene skeleton (ω₁ = +63°, ω₃ = −40°) and an antiperiplanar orientation of the carbonyl and methoxy groups at the pendants. The theoretical results (Fig. 6b and see ESI^† for additional information) match with the experimental data, indicating that our model structure is a good approximation to the helical structure adopted by poly-(S)-3.Finally, the stimuli response properties of poly-(S)-3 were explored by CD. These experiments revealed that the addition of monovalent or divalent metal ions to a chloroform solution of poly-(S)-3 does not produce any significant effect in the structural equilibrium of this polymer (see Fig. S18^†). This fact, in addition to the previous results obtained from the interaction of poly-(S)-2 with divalent metal ions, corroborates the decrease of the dynamic character of helical PPAs when large OPEs are used as spacers.The poor dynamic behaviour was further demonstrated by polymerizing m-(S)-3 at a lower temperature (0 °C) (Fig. 6d). In this case, the region around 240–350 nm remains unaffected, indicating that the pendant is ordered in a similar manner in both batches of polymers, regardless of the temperature at which they were synthesized (20 °C and 0 °C). Interestingly, the magnitude of the first Cotton band is duplicated when the polymer is obtained at low temperature due to a stronger helical sense induction at the polyene backbone. This result indicates that a preorganization process may occur during polymerization, affecting the screw sense excess of the PPA.In conclusion, a novel chiral harvesting transmission mechanism has been described in poly(acetylene)s bearing oligo(p-phenylenethynylene)s as rigid spacers that place the chiral pendant group away from the polyene backbone, at a distance around ca. 1.7 nm for poly-2, and 2.4 nm for poly-3. Hence, the disposition of the chiral moiety affects the stacking of the OPE units within the helical structure, inducing a specific positive or negative tilting degree, which is further harvested by the polyene backbone inducing either a P or M internal helix.We believe that these results open new horizons in the development of novel helical structures by combining information from the helical polymers and supramolecular helical polymers fields, which leads to the formation of novel materials with applications in important fields such as asymmetric synthesis, chiral recognition or chiral stationary phases among others.     

Hyperpolarisation of weakly binding N-heterocycles using signal amplification by reversible exchange

Signal Amplification by Reversible Exchange (SABRE) is a catalytic method for improving the detection of molecules by magnetic resonance spectroscopy. It achieves this by simultaneously binding the target substrate (sub) and para-hydrogen to a metal centre. To date, sterically large substrates are relatively inaccessible to SABRE due to their weak binding leading to catalyst destabilisation. We overcome this problem here through a simple co-ligand strategy that allows the hyperpolarisation of a range of weakly binding and sterically encumbered N-heterocycles. The resulting ¹H NMR signal size is increased by up to 1400 times relative to their more usual Boltzmann controlled levels at 400 MHz. Hence, a significant reduction in scan time is achieved. The SABRE catalyst in these systems takes the form [IrX(H)₂(NHC)(sulfoxide)(sub)] where X = Cl, Br or I. These complexes are shown to undergo very rapid ligand exchange and lower temperatures dramatically improve the efficiency of these SABRE catalysts.

The scope of the hyperpolarisation method Signal Amplification by Reversible Exchange (SABRE) is dramatically expanded through the use of co-ligands to substrates that weakly interact with the active cataylst.

Hyperpolarised magnetic resonance is receiving increasing attention from both the analytical science and medical communities due to its ability to create signals that are many orders of magnitude higher than those normally detected under Boltzmann control.^1–6 The time and cost benefits associated with this improvement have propelled this area of research forward over the past few decades. Two of the most prominent techniques used to create hyperpolarisation are dissolution Dynamic Nuclear Polarisation (d-DNP) and Para-Hydrogen Induced Polarisation (PHIP),^7,8 which derive their non-Boltzmann spin energy level populations from interactions with unpaired electrons and para-hydrogen (p-H₂, the singlet spin isomer of hydrogen), respectively. Both of these methods have been reviewed in detail.^3–5,9,10Signal Amplification by Reversible Exchange (SABRE) is a PHIP method that does not involve the chemical incorporation of p-H₂ into the target substrate.^11,12 Instead, under SABRE, spin order transfer proceeds catalytically through the temporary formation of a scalar coupling network between p-H₂ derived hydride ligands and the substrate''s nuclei whilst they are located in a transient metal complex. The most common catalysts are of the type [Ir(H)₂(NHC)(sub)₃]Cl (where NHC = N-heterocyclic carbene and sub = the substrate of interest, Fig. 1a),^13,14 although other variants are known.^15–17 For SABRE to be accomplished, the target substrate must be able to reversibly ligate to the metal centre and this limits the methods applicability; although several routes to overcome this have been reported.^18–20 Recently, the use of bidentate ancillary ligands such as NHC-phenolates¹⁶ and phosphine-oxazoles²¹ has been shown to expand the applicability of SABRE for a variety of different ligands and solvents (Fig. 1b). For example, use of the PHOX ligand (PHOX = (2-diphenylphosphanyl)phenyl-4,5-dihydrooxazole) gives ¹H NMR signal gains of up to 132-fold for 2-picoline; a substrate previously shown to be unpolarised under classic SABRE conditions.²²Open in a separate windowFig. 1Development of the SABRE method for hyperpolarisation of a range of substrates.The use of co-ligands to stabilise the active SABRE catalyst has proven successful for substrates that weakly associate to the catalyst (Fig. 1c). Of particular note is the hyperpolarisation of sodium [1,2]-¹³C₂-pyruvate²³ and sodium ¹³C-acetate²⁴ which could be used as in vivo metabolic probes. The importance of co-ligands in breaking the chemical symmetry of the SABRE catalyst is also well established and co-ligands such as acetonitrile,²⁵ sulfoxides,^23,26 1-methyl-1,2,3-triazole²⁷ and substrate isotopologues²⁸ have been employed.We report here on the use of co-ligands to allow the NMR hyperpolarisation of weakly binding N-heterocyclic derived substrates with functionality in the ortho-position that have proven to be routinely inaccessible to the SABRE technique (Fig. 1d). ¹H signal gains of up to 1442 ± 84-fold were obtained for some of these substituted pyridines at 9.4 T and the expansion of this approach to ¹³C and ¹⁵N detection and other N-heterocyclic motifs is also exemplified.     

Heparin reversal by an oligoethylene glycol functionalized guanidinocalixarene

Unfractionated heparin (UFH), a naturally occurring anionic polysaccharide, is widely used as an anticoagulant agent in clinical practice. When overdosed or used in sensitive patients, UFH may cause various risks and a UFH neutralizer needs to be administered immediately to reverse heparinization. However, the most common UFH neutralizer, protamine sulfate, often causes various adverse effects, some of which are life-threatening. Herein, we designed a highly biocompatible, oligoethylene glycol functionalized guanidinocalixarene (GC4AOEG) as an antidote against UFH. GC4AOEG and UFH exhibited a strong binding affinity, ensuring specific recognition and neutralization of UFH by GC4AOEG in vitro and in vivo. As a consequence, UFH-induced excessive bleeding was significantly alleviated by GC4AOEG in different mouse bleeding models. Additionally, no adverse effects were observed during these treatments in vivo. Taken together, GC4AOEG, as a strategically designed, biocompatible artificial receptor with strong recognition affinity towards UFH, may have significant clinical potential as an alternative UFH reversal agent.

An oligoethylene glycol functionalized guanidinocalix[4]arene was developed as a safe antidote against heparin, via specific recognition and neutralization of heparin in vitro and in vivo.

Heparin sodium, often referred to as unfractionated heparin (UFH, also known as heparin), is a well-known anionic glycosaminoglycan consisting of long, helical, unbranched chains of repeating sulfonated disaccharide units (Fig. 1).¹ It is currently a gold-standard life-saving drug to overcome blood coagulation by activating antithrombin-III to impede the coagulation process.^2,3 Systemic heparinization is the most common anticoagulation procedure in surgical practice (e.g. open-heart surgery) and extracorporeal therapies such as kidney dialysis. At the end of surgery, excess heparin often needs to be deactivated by using a heparin neutralizer; otherwise patients have risks of low blood pressure and a slow heart rate, and may develop internal bleeding.⁴ Therefore, the neutralization of heparin has been a topic of significant research interest in the biomedical field.Open in a separate windowFig. 1Scheme of heparin reversal by GC4AOEG in the circulatory system.Protamine sulfate, the only FDA-approved neutralizer of UFH, possesses a highly positive charge density due to its polymeric nature and rich arginine residues. Thus, electrostatic interactions are the major driving force in the formation of a UFH–protamine complex, leading to the neutralization and deactivation of UFH.^1,5 However, due to its non-specific interactions, protamine sulfate often causes various adverse effects such as bradycardia, hypotension and pulmonary hypertension, as well as allergic reactions including life-threatening anaphylactic reactions in some patients.⁵ When overdosed, protamine may further impair the intricate balance in the blood and cause coagulopathy.^5–7 Given these issues, there has been a medical need for alternative, safe UFH neutralizers that can specifically counteract UFH without causing serious adverse effects.⁸Discovering and developing new heparin neutralizers has been a popular area of research.^8,9 During the past two decades, a variety of different UFH neutralizers including small molecules,¹⁰ cationic polymers (e.g. polybrene),^11–14 peptides,^11,15 and nanoparticles^16,17 have been designed and evaluated in vitro and/or in vivo. For instance, surfen, as a small-molecule antagonist of UFH, may electrostatically bind with UFH; however only modest neutralizing effects against UFH were observed in rats,^10,18 likely attributed to the lack of strong, specific recognition. On the other hand, polycationic species, including polybrene¹⁹ and poly-dl-lysine,²⁰ exhibited stronger binding with UFH and significant potential as UFH neutralization agents. However, toxicity was still a key concern of these species due to their intrinsic electrostatic interactions with red blood cells (RBC).²¹ Meanwhile, some UFH antagonists have achieved preliminary success in preclinical studies and even moved to clinical evaluations. For instance, ciraparantag (PER977), as a synthetic antidote against several anticoagulants, is currently being evaluated in phase II clinical trials.²² UHRA (Universal Heparin Reversal Agent), a synthetic multivalent dendrimer polymer in the form of nanoparticles with positively charged surfaces, can reverse the activity of all clinically available heparins and it is currently undergoing preclinical studies and will likely move to clinical investigations.²³ However, the oligo- and poly-cationic nature of these species suggests their general tendency towards any negatively charged species, making them “universal” or function against several anticoagulants, implying their low specificity towards heparin.More recently, the sequestration and reversal of toxic agents by supramolecular host molecules have attracted increasing attention, and a typical example of clinical and commercial success is sugammadex, a carboxylated derivative of gamma-cyclodextrin that may specifically reverse the activity of non-depolarizing neuromuscular blocking agents.²⁴ Inspired by this clinical success, several macrocycles were designed and synthesized to selectively bind UFH. For instance, Liu et al. synthesized amphiphilic multi-charged cyclodextrins (AMCD), and AMCD-assembly was utilized for selective heparin binding.¹⁶ Nitz et al. derivatized a cyclodextrin with amide and guanidino groups as a polycationic receptor to recognize and detect UFH.²⁵ Kostiainen and co-workers studied cationic, quaternary ammonium functionalized pillar[5]arene because of its potential complexation with UFH.²⁶ Additionally, cationic calixarene derivatives were designed for UFH binding and guanidinocalixarenes exhibited stronger binding affinity with UFH than their quaternary amine-functionalized counterparts.^27,28 In spite of decent binding affinities and selective recognition of UFH, these macrocycles still possess various limitations such as non-specific toxicity induced mostly by cationic charges, which may disrupt cell membranes and induce blood coagulation.^29,30An ideal UFH neutralizer should full-fill the following three requirements: (1) binding strongly towards UFH in a specific manner; (2) excellent biocompatibility and safety profile, and (3) a clearly defined molecular structure to facilitate batch-to-batch consistency. Thus far, none of the clinical UFH antagonists or previously reported candidates has fulfilled these conditions. Herein we designed an artificial receptor, an oligoethylene glycol functionalized guanidinocalixarene, GC4AOEG, by leveraging the asymmetrical structure of calixarene to strategically add guanidinium groups on one side and oligoethylene glycol (OEG) groups on the other side (Fig. 1). We anticipated that the guanidinium-enriched upper rim would bind strongly with UFH via salt bridges (charge-assisted hydrogen bonds).^28,31 In addition, the biocompatible OEG-functionalized lower rim may help improve the water-solubility and biocompatibility of the host molecule.^32,33GC4AOEG was synthesized in 5 steps starting from the maternal calix[4]arene (Fig. 2). Briefly, p-tert-butylcalix[4]arene 1 was alkylated with tosylate 2³⁴ to obtain compound 3 with a well-defined cone conformation, and replacement of the tert-butyl with nitro groups via an ipso-nitration reaction afforded compound 4.³⁵ Subsequently, compound 4 was hydrogenated in the presence of SnCl₂·2H₂O, affording the tetramine derivative 5. Subsequently, compound 6 was obtained via a reaction between compound 5 and di-Boc-protected thiourea units. The removal of the protecting groups was achieved using SnCl₄ in ethyl acetate, to yield the target GC4AOEG (the characterization of intermediates (Fig. S1 and S2) and GC4AOEG (Fig. S3) are in the ESI^†).Open in a separate windowFig. 2Synthetic route of GC4AOEG and fluorescence titrations. (A(a)) NaH, dry DMF, and 75 °C; (b) HNO₃, AcOH, dry CH₂Cl₂, and r.t.; (c) SnCl₂·2H₂O, C₂H₅OH/AcOEt (1 : 1, v/v), and reflux; (d) 1,3-bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea, Et₃N, AgNO₃, dry CH₂Cl₂, and r.t.; (e) SnCl₄, AcOEt, and r.t. (B) Direct fluorescence titration of 0.5 μM EY with different concentrations of GC4AOEG (up to 13.8 μM) in HEPES buffer (10 mM, pH = 7.4), and λ_ex = 517 nm. (Inset) The associated titration curve at λ_em = 537 nm and best fit according to a 1 : 1 binding stoichiometry. (C) Competitive fluorescence titration of GC4AOEG·EY (4.0/0.5 μM) with UFH (up to 8.4 μM in the concentration of monomer units of UFH), and λ_ex = 517 nm. (Inset) The associated titration curve at λ_em = 537 nm and best fit according to a n : 1 competitive binding model, where n = 0.88.The binding affinity between GC4AOEG and UFH was firstly investigated via a competitive titration approach. In this paper, we defined the repeated disaccharide unit as the UFH monomer unit, and the UFH concentration in this paper is the UFH monomer unit concentration. Eosin Y (EY) was selected as the reporter dye, owing to its strong complexation with GC4AOEG and the drastic fluorescence quenching after complexation. The equilibrium association constant (K_a), between GC4AOEG and EY, was determined by direct fluorescence titration and fitted as (2.37 ± 0.12) × 10⁵ M⁻¹ with 1 : 1 binding stoichiometry (Fig. 2B). The displacement of EY from GC4AOEG·EY by gradual addition of UFH resulted in the recovery of the intrinsic emission of EY. The best-fitting of the competitive titration model afforded ca. 1 : 1 binding stoichiometry between GC4AOEG and each monomer unit of UFH, as well as an ultrahigh binding affinity K_a of (1.25 ± 0.13) × 10⁷ M⁻¹ (Fig. 2C).For in vitro analysis of the effectiveness of GC4AOEG against UFH, the activated partial thromboplastin time (aPTT) assay was conducted. The result (Fig. S8^†) indicates that one equivalent of GC4AOEG (to UFH monomer) fully neutralized UFH, similar to protamine. Very importantly, it is obvious that protamine alone negatively influenced the aPTT time. In contrast, GC4AOEG alone did not affect the clotting time, suggesting that GC4AOEG can specifically bind with UFH directly with minimal side influences. The coagulation factor X levels in the plasma analyzed via the enzyme-linked immunosorbent assay (ELISA) further confirmed the safety and reversal effect of GC4AOEG towards UFH (Fig. S9^†).Next, the biocompatibility of GC4AOEG was investigated in vitro. As an alkyl derivative of guanidinocalixarene, GC4A-6C (Fig. S4 and S5^†), which has a similar number of carbons (hexyl groups) at the lower rim to that of GC4AOEG, was also synthesized and examined in this study for comparative purposes. As shown in Fig. 3A and B, GC4AOEG (up to 200 μM) showed remarkably low cytotoxicity in several cell lines via MTT assays, in dramatic contrast to the relatively high cytotoxicity of GC4A-6C (Fig. 3C and D). The cellular toxicity of GC4A-6C was consistent with previous literature.³⁶ In addition, alkyl derivatives of calixarene were generally more toxic than those without alkyl chains,³⁷ likely attributed to their amphiphilic properties that may facilitate cell membrane disruption.^38–40 The results suggested that the much-improved safety profile of GC4AOEG was attributed to oligoethylene glycol functionalization. Meanwhile, it is well known that cationic polymers or oligomers often show poor biocompatibility in the circulation system due to their non-selective binding to negatively charged RBC, resulting in RBC aggregation or hemolysis.⁴¹ Therefore, hemolysis and hemagglutination assays were conducted according to a method previously reported,^42,43 with experimental details described in the method. The percent hemolysis of GC4A-6C (25, 50, 100 and 200 μM, respectively) was over 90%, which would limit its application in the circulatory system (Fig. S6^†), as a hemolysis ratio below 5% is considered safe.⁴⁴ Conversely, GC4AOEG exhibited nearly negligible (less than 3%) hemolytic activity at concentrations of up to 200 μM, and no agglutination was visualized during incubation with RBC (Fig. 3F), implying that OEG functionalization at the lower rim reduced non-specific interactions with the RBC membrane, resulting in less disturbance of the membrane structure and function or cellular aggregations.Open in a separate windowFig. 3Biocompatibility study in cell lines and RBC. Cell viabilities of (A, C) 4T1 and (B, D) 293T, cells treated with different concentrations of GC4AOEG or GC4A-6C for 24 h. Each data point represents the mean ± S.E.M. from a set of experiments (n = 4). (E, G) Hemolysis test of GC4AOEG at different concentrations (NC = negative control; PC = positive control). Each data point represents the mean ± S.E.M. from a set of experiments (n = 3). (F) Agglutination test of RBC incubated with GC4AOEG at 2.0% hematocrit in normal saline.Inspired by the above findings, we further examined whether GC4AOEG may reverse bleeding in different mouse bleeding models under heparinization (with the experimental details described in the method, and the standard curve for the quantification of blood loss volume is showed in Fig. S7^†),⁴⁵ with both the total time of bleeding and total volume of lost blood evaluated for each model. As a proof of concept, 200 U kg⁻¹ UFH and 2.245 mg kg⁻¹ GC4AOEG (molar ratio of GC4AOEG and each monomer unit of UFH = 1 : 1) were respectively used, as representative doses in the study and the dose of UFH was based on a literature report.⁴⁶ In a mouse tail transection model as an external bleeding model, as shown in Fig. 4A–C, after tail transection, the bleeding time and blood loss volume for mice treated with normal saline were 58.9 ± 10.7 min and 72.2 ± 15.8 μL, respectively. As expected, treatment with UFH increased the bleeding time and blood volume to 121.5 ± 20.2 min and 264.0 ± 43.6 μL, respectively. In contrast, the bleeding time was dramatically reduced down to the blank control level, when the mice were treated with GC4AOEG at the same time of, or 30 s after, i.v. administration of UFH (53.8 ± 11.4 min and 89.0 ± 13.3 min, respectively). Accordingly, the blood loss volume of mice successively treated with UFH and GC4AOEG (1 : 1 ratio) reached the control level (72.6 ± 14.3 μL), indicating that the strong binding affinity between GC4AOEG and UFH ensured their recognition in vivo. Of note, there was no significant difference between the GC4AOEG treated group (without heparinization) and the saline treated group, suggesting a decent safety profile of the artificial receptor.Open in a separate windowFig. 4Reversal efficacy in in vivo mouse models. (A–C) Mouse tail transection model. (A) Scheme of the mouse tail transection model. (B) Total time of bleeding and (C) blood loss volume. (D–F & J) Mouse liver injury model. (D) Scheme of the mouse liver injury model. (E) Total time of bleeding and (F) blood loss weight. (J) Pictures exhibiting bleeding in liver injury before and after treatment. (G–I & K) Mouse femoral artery model. (G) Scheme of the mouse femoral artery model. (H) Total time of bleeding and (I) blood loss weight. (K) Pictures exhibiting bleeding in the femoral artery before and after treatment. All of those models were i.v. administration with normal saline (control), GC4AOEG (2.245 mg kg⁻¹), or UFH (200 U kg⁻¹) without and with GC4AOEG (2.245 mg kg⁻¹, 1 : 1 molar stoichiometry of GC4AOEG and the monomer unit of UFH), and UFH–GC4AOEG 1 : 1 successively (GC4AOEG at a dose of 2.245 mg kg⁻¹ 30 s after UFH administration) respectively were quantified. Data presented are the mean ± S.E.M. (n = 6). *p < 0.05, ****p < 0.001, and ns represents “no significant difference” between the experimental group and the control group.In addition to external bleeding, internal bleeding such as liver injury model (Fig. 4D) was established in mice, and GC4AOEG''s reversal of UFH was further evaluated in vivo. Mice were i.v. administered with normal saline (control), GC4AOEG (2.245 mg kg⁻¹), or UFH (200 U kg⁻¹) without and with GC4AOEG (2.245 mg kg⁻¹), and successive UFH–GC4AOEG 1 : 1 (30 s in between), respectively. In 2 minutes, the abdomen was surgically opened to expose the liver. A wound of 0.5 cm length and 2 mm depth, in the left lobe of the liver, was created. Considerable bleeding was immediately observed in the UFH treatment group (Fig. 4J), with the total bleeding time lasting for 450.5 ± 46.8 s, and the total blood loss of 571.0 ± 35.0 mg, in contrast to 143.7 ± 14.7 s total bleeding time and 238.0 ± 45.0 mg total blood loss observed in the saline treated group. Interestingly, the UFH–GC4AOEG treated group showed no significant difference from the normal saline treated group. To simulate the clinical use scenario, GC4AOEG was injected after UFH''s administration, and significantly reduced bleeding (from both time and volume perspectives) was observed, suggesting effective inhibition of the adverse effects of UFH, by GC4AOEG (Fig. 4E and F). GC4AOEG alone (without heparinization) did not exhibit any hematological toxicity in this model. To further evaluate the inhibitory effects of GC4AOEG against UFH in a preclinical model, a more serious internal bleeding model, femoral artery bleeding mouse model, was employed, and the treatment plan followed the previous two models described as above. Upon administration, the skin of the right leg and the overlying muscles were removed to expose the femoral artery and sciatic nerve. After an open injury at the middle segment of the femoral artery was created with a surgical scissor, blood gushed out immediately from the injured site (Fig. 4G and K). As shown in Fig. 4H and I, the longest average bleeding time (16.0 ± 1.9 min) and blood loss weight (103.8 ± 16.9 mg) were observed in the UFH treatment group of mice, in dramatic contrast to the bleeding time and blood loss of 3.9 ± 0.4 min and 24.7 ± 4.5 mg, respectively, in the normal saline treated group of mice. A bleeding time of 3.6 ± 0.4 min and blood loss of 20.8 ± 7.4 mg were recorded in the UFH–GC4AOEG treatment group. When UFH and GC4AOEG (at 1 molar equivalent) were successively injected, a bleeding time of 5.3 ± 0.7 min and blood loss of 27.7 ± 5.8 mg were noted, suggesting the significant reversal effects of GC4AOEG on UFH. Collectively, in all of the three bleeding models including internal and external bleeding models, i.v. administration of GC4AOEG significantly reversed UFH-induced excessive bleeding in external and internal injuries. More importantly, GC4AOEG alone exhibited negligible hematological activity, unlike other previously reported cationic small molecules, polymers, oligomers and macrocycles.Furthermore, in order to further verify the safety profile of GC4AOEG at the effective dose in vivo, acute toxicity evaluation was performed in a mouse model. After the i.v. injection of GC4AOEG in mice at a dose of 2.245 mg kg⁻¹ (i.v. injection of normal saline as the control group), the body weight, behaviors, and overall survival of the treated mice were monitored every day for 3 weeks. All the treated mice remained alive and showed normal behaviors, as well as normal body weight evolvement similar to that of the control group (Fig. 5A). On day 21 post administration, mice were euthanized for blood and organ samples were harvested (for details see the method). The organ indexes of representative major organs including the heart, liver, spleen, lungs, and kidneys isolated from the GC4AOEG treated mice were comparable to those of the mice administered with normal saline, with no significant differences observed (Fig. 5B). Hematological parameters such as the counts of whole blood cells (WBCs), red blood cells (RBCs), platelets (PLTs) and hemoglobin (HGB) (Fig. 5C), as well as the serum concentrations of liver and kidney function biomarkers including blood urea nitrogen (BUN), creatinine (crea), urea alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were all analyzed thoroughly (Fig. 5D and E). These results indicated that the hematological parameters, renal and hepatic functions of the mice treated with GC4AOEG were comparable with those of the mice in the normal saline treated group. Moreover, histopathological examinations of the major organs of the GC4AOEG treated mice showed normal microstructures comparable with those of the control group (Fig. 5F). Collectively, these results suggested that the i.v. administration of GC4AOEG at the therapeutic dose is safe.Open in a separate windowFig. 5Preliminary acute toxicity evaluations on GC4AOEG. (A) Weight changes of mice after i.v. administration with a single dose of GC4AOEG. (B) Major organ indexes of the mice on day 21 post-administration with GC4AOEG. (C) Hematological parameters of the blood samples collected from the mice on day 21 after i.v. administration of GC4AOEG. (D) Renal and (E) hepatic functional biomarkers in the blood samples collected from the mice on day 21 after i.v. administration of GC4AOEG. Data are presented as mean ± S.E.M.; n = 6 for each group. (F) H&E histopathological analysis of the major organs from mice sacrificed 21 days after being injected with saline and GC4AOEG (2.245 mg kg⁻¹). Scale bar = 100 μm.     

Boron tribromide as a reagent for anti-Markovnikov addition of HBr to cyclopropanes

Although radical formation from a trialkylborane is well documented, the analogous reaction mode is unknown for trihaloboranes. We have discovered the generation of bromine radicals from boron tribromide and simple proton sources, such as water or tert-butanol, under open-flask conditions. Cyclopropanes bearing a variety of substituents were hydro- and deuterio-brominated to furnish anti-Markovnikov products in a highly regioselective fashion. NMR mechanistic studies and DFT calculations point to a radical pathway instead of the conventional ionic mechanism expected for BBr₃.

Anti-Markovnikov hydrobromination of cyclopropanes was achieved using boron tribromide and water as the bromine and proton sources, respectively.

The Lewis acidic nature of organoboranes is well understood, but the participation of BR₃ in free-radical processes was largely overlooked until 1966.¹ Since the discovery of the potential of organoborane species to undergo radical reactions, many novel and synthetically useful transformations were developed.² Trialkylboranes (BR₃) can easily undergo bimolecular homolytic substitution (S_H2) at the boron atom to generate alkyl radicals (Scheme 1A). It was found that alkoxyl, dialkylaminyl, alkylthiyl and carbon-centered radicals, triplet ketones, and triplet oxygen can all initiate the radical reaction by substituting one of the alkyl groups of trialkylboranes to liberate alkyl radicals.³ BEt₃/O₂ is arguably the most studied organoborane radical-initiating system, with the peroxyl radical being the key to propagate the reaction. Apart from being a radical initiator, BEt₃, along with trace amount of O₂, can also undergo conjugate addition to unsaturated ketones and aldehydes; addition to ethenyl- and ethynyloxiranes, azidoalkenes, and imines; and addition–elimination to nitroalkenes and nitroarenes, styryl sulfones, sulfoxides and sulfinimides.³ However, apart from changing BEt₃ to other trialkylboranes or catecholborane to carry out similar radical reactions, the radical-reaction potential of other organoboranes remains underexplored, given the ease and mild conditions under which they initiate radical chains, often with trace amount of O₂ in air at low temperature. The application of such a mild radical-initiation system to stereoselective radical reactions would drastically change the reaction outcome especially when intermediates and products are thermally unstable.⁴Open in a separate windowScheme 1Classical radical reactions with trialkylboranes and our work on radical bromination using BBr₃.Halogenation is an important class of transformations and the resultant halogenated products can easily be manipulated to give a wide range of functional molecules.⁵ While trihaloboranes have been employed as halogenating or haloborating agents, their role in reactions is either ambiguous or thought to be exclusively Lewis acidic.⁶ To date, the use of trihaloboranes as a halogen radical donor has not been reported. With BR₃/O₂ being a versatile radical-initiator and conjugate-addition system, we envisioned that a suitable halogenated-borane might work similar to that of trialkylboranes in the generation of reactive, yet stable enough halogen radicals for selective halogenation reactions (Scheme 1B).Trialkylboranes readily undergo S_H2 reactions because the formation of stronger B–X (e.g. B–O) bonds via substitution is highly exothermic.³ The BDEs (B–C) of BMe₃, BEt₃, BⁿPr₃, BⁱPr₃, and BⁿBu₃ range from 344 to 354 kJ mol⁻¹ at 298 K, while their typical autoxidation products, B(OH)₃, B(OMe)₃, and B(OEt)₃, have BDEs (B–O) ranging from 519 to 522 kJ mol⁻¹ at 298 K.⁷ We hypothesized that organohaloboranes (BX_aR_3−a, X = halogen) with BDEs (B–X) similar to trialkylboranes would be a halogen radical donor from a thermodynamic viewpoint. As the common trihaloboranes (BX₃) BF₃, BCl₃ and BBr₃ have BDEs (B–X) of 644.3, 442.3 and 367.1 kJ mol⁻¹ at 298 K, respectively, BBr₃ was the logical option for our purpose.⁸ Although the BDE (B–I) of BI₃ is the lowest among all trihaloboranes and found to be 278.2 kJ mol⁻¹ at 0 K,⁹ it was not considered suitable as I₂ has proven to be a very efficient radical quencher in such reactions,¹⁰ and even rigorously purified BI₃ invariably contains a trace amount of I₂.¹¹Compared to activated cyclopropanes,¹² oxidative functionalization of unactivated cyclopropanes gives a wide range of useful molecules that are otherwise not readily accessible, and protocols for the Markovnikov-selective functionalization of unactivated cyclopropanes have been reported.^13–20 Halolyses of cyclopropanes to give 1,3-dihaloalkanes by molecular halogens are also documented although the reactions commonly suffer from the formation of side products via electrophilic aromatic halogenation.²¹ In contrast, obtaining products with anti-Markovnikov regioselectivity has been considered as one of the top challenges in industry.^22–30 Anti-Markovnikov functionalization of unactivated cyclopropanes mostly relies on photo-initiated radical processes with generally poor regioselectivity and limited scope.^31–36 To the best of our knowledge, anti-Markovnikov hydrohalogenation of cyclopropanes has not been reported.Very recently, an anti-Markovnikov hydroboration for unactivated cyclopropanes has been reported using boron tribromide and phenylsilane.³⁷ The reaction was carried out under inert and anhydrous conditions, and mechanistic studies pointed to an ionic mechanism with Lewis acid–base interactions. We show that with a simple twist in the reaction conditions, which is to introduce oxygen, a drastically different reaction outcome and mechanism could be realized. We now report the study and application of BBr₃ as a radical Br donor for the anti-Markovnikov addition of HBr to cyclopropanes.With all these considerations in mind, we initially envisioned that BBr₃/O₂ as a suitable system to generate bromine radicals, and cyclopropylbenzene (1a) as the model substrate to capture them. The radical reaction might then be terminated by another halogen radical from reagents such as N-chlorosuccinimide or N-iodosuccinimide. Unfortunately, messy mixtures were obtained for all entries (see the ESI, Scheme S1^†). On the other hand, a simple proton source, H₂O, was found to be effective in terminating the radical species. In the control experiment with only BBr₃ and cyclopropylbenzene (1a) (Scheme 2, entry 1), the anti-Markovnikov hydrobrominated product 2a was obtained in 24% yield, together with the formation of Markovnikov product 3a (trace) and dibrominated cyclopropane 4a (11%). We reasoned that the proton source was the trace amount of moisture in commercial BBr₃ solution.Open in a separate windowScheme 2Reaction optimization. Conditions: reactions were carried out under ambient conditions and quenched by saturated NaHCO₃ solution. Yields were measured by ¹H NMR with CH₂Br₂ as the internal standard. ^a24% of 1a was recovered. ^b6% of 1a was recovered.Although it is well known that boron-based Lewis acids are moisture sensitive,³⁸ counter-intuitively, the addition of 1.5 equivalents of H₂O had a positive impact on the yield of 2a, which was dramatically improved to 80% (Scheme 2, entry 2). Excess water led to a reduction in the yield of 2a and the regioselectivity (Scheme 2, entry 3). Replacing water with ethanol as the proton source resulted in a significant drop in reaction efficiency (Scheme 2, entry 4). In contrast, bulkier alcohols such as i-PrOH or t-BuOH (Scheme 2, entries 5 and 6) and less nucleophilic alcohols such as CF₃CH₂OH (Scheme 2, entry 7) gave comparable performance to that of water.Further study revealed that achieving anti-Markovnikov addition of HBr to cyclopropanes in conventional systems is not a trivial task (Scheme 3). For instance, no reaction was observed when 1a was treated with HBr in either aqueous or water/AcOH co-solvent systems at room temperature.²⁹ Heating both reactions only yielded the Markovnikov product 3a in 16–23% yield, and no anti-Markovnikov product 2a was detected. The classical radical bromination protocol with BBr₃/H₂O₂ only furnished dibrominated product 4a in 29% yield. Similar to the uniqueness of BR₃/O₂ in several aforementioned radical reactions,⁴ the incapability of these control experiments in producing 2a as a product contrasted starkly with our BBr₃/O₂ conditions, which generated a reactive yet selective bromine radical.Open in a separate windowScheme 3Reactions of cyclopropane (1a) with hydrobromic acid.Next, we expanded the substrate scope to other unactivated cyclopropanes using either water or t-BuOH as the proton source (Scheme 4). Electron-neutral, deficient and sterically bulky substrates 1a–1g gave the desired anti-Markovnikov products 2a–2g in good yields and regioselectivity. Cyclopropanes with electron-deficient substituents including nitriles (1j–n) and ester (1o) also worked well with excellent regioselectivity. This protocol also exhibits high chemoselectivity towards cyclopropanes. Aryl methyl ether (2i), which is known to be easily cleaved by BBr₃ even at low temperature, remained intact under our reaction conditions.³⁹ Due to the tendency of aryl vinyl ketones to polymerize, they are known to be unsuitable for 1,4-conjugate additions mediated by trialkylboranes.⁴⁰ Nevertheless, aryl cyclopropyl ketones (1h–i) were converted into the corresponding products in high yields, and polymerization was not observed. 1,1-Disubstituted (1p) and simple alkyl (1q) cyclopropanes were also compatible to give products 2p and 2q. When cyclopropyl carboxylic acid (1r) was used as the substrate, the unstable product 2r was detected using HRMS and crude ¹H NMR, and γ-butyrolactone was obtained ultimately through cyclization upon a basic work-up procedure. Indene-derived cyclopropyl substrate 1s was also compatible to give 2s. Scaled-up reactions were also performed on selected examples (2a, 2h, 2o, and 2r) and excellent regioselectivities were still obtained.Open in a separate windowScheme 4Reaction scope of anti-Markovnikov hydrobromination of cyclopropanes. Conditions: reactions were carried out with 1 (0.2 mmol) unless stated otherwise. Exact reaction conditions for each substrate are stated in the ESI.^†at-BuOH was used as the proton source. ^bH₂O was used as the proton source. ^c4% of 3b was detected. ^d5% of 3c was detected. ^e7% of 3b was detected. ^fThe reaction was conducted on a 1 mmol scale. ^gThe reaction was conducted on a 2 mmol scale. ^hThe product cyclized quickly upon work-up and the yield was measured on the basis of the cyclized product γ-butyrolactone.Cyclopropanes 1t–1y with secondary and tertiary alcohols also gave the corresponding anti-Markovnikov products in excellent yields and with high regioselectivities (Scheme 5). The structure of 2x was confirmed unambiguously by X-ray crystallography.⁴¹ The hydroxyl groups in the substrates were converted into bromides simultaneously by the action of BBr₃ to give a series of useful dibromides.⁴² We were interested in whether alcohol-containing substrates can be hydrobrominated in the absence of an external proton source. To our delight, 1t was able to undergo anti-Markovnikov hydrobromination to give 2t with only a slight drop in yield (76%), and 2u was produced in quantitative yield. The hydroxyl groups in 1x and 1y appear to be crucial because a sluggish reaction was observed for 1-phenyl-2-methylcyclopropane that bears no hydroxyl substituent.Open in a separate windowScheme 5Reaction scope of anti-Markovnikov hydrobromination of cyclopropanes with hydroxyl substituents. Conditions: reactions were carried out with 1 (0.2 mmol). Exact reaction conditions for each substrate are stated in the ESI.^†aReaction was conducted in the absence of water. ^bDiastereoselectivity was determined by a ¹H NMR experiment on the crude mixture.By substituting H₂O and t-BuOH with D₂O and t-BuOD, deuteriobrominations were also carried out and the corresponding mono-deuterium-labeled compounds were obtained smoothly (Scheme 6). Our protocol offered excellent regio-control in the mono-deuteriation to give 2-D. Unactivated (1a and 1e) and activated cyclopropanes (1j–k, 1m, and 1o) with various substituents worked well and excellent levels of deuterium incorporation were achieved.Open in a separate windowScheme 6Reaction scope of anti-Markovnikov deuteriobromination of cyclopropanes. Conditions: reactions were carried out with 1 (0.2 mmol) unless stated otherwise. Exact reaction conditions for each substrate are stated in the ESI.^†aThe % D incorporation was determined based on the integration of the residual proton signal in ¹H NMR. ^bThe reaction was conducted on a 1 mmol scale. ^ct-BuOH was used as the deuterium source. ^dD₂O was used as the deuterium source.The conversion of products 2 into primary alcohols and amines through nucleophilic substitution proved straightforward. For instance, alcohol 5a and amine 5b were readily prepared from 2a with high conversion (Scheme 7). As the direct synthesis of primary alcohols and amines through anti-Markovnikov hydration and hydroamination has proven to be challenging,²² our protocol provides useful precursors for the synthesis of these highly desired compounds.Open in a separate windowScheme 7Synthetic utilities of 2a.We envision a radical reaction pathway between BBr₃ and O₂, but given the Lewis acidity of BBr₃ and Lewis basicity of H₂O and alcohols, an acid-mediated pathway cannot be ruled out.³⁸ However, such a pathway appears highly unlikely, as the treatment of cyclopropanes with aqueous HBr yielded no anti-Markovnikov product 2 (Scheme 3). Several control experiments were performed to further probe the reaction mechanism.The addition of a radical scavenger, BHT or TEMPO, in slight excess of BBr₃ completely shut down the formation of anti-Markovnikov product 2a, and a significant amount of Markovnikov product 3a was detected (Scheme 8A). The addition of the acceptor olefin acrylonitrile completely suppressed the reaction. The absence of light had no impact on the reaction, thereby eliminating the possibility of a photo-triggered pathway. The presence of oxygen was crucial for both the yield and the regioselectivity. The reaction proceeded smoothly to give the desired product 2a (80%) in open air. In contrast, the yield of anti-Markovnikov product 2a dropped to 14% and that of the Markovnikov product 3a increased to 17% when the reaction was conducted with degassed CH₂Cl₂ and 1a. Deuteriobromination of 1g was also conducted with t-BuOD as the deuterium source (Scheme 8B). Other than the benzylic deuteriation product 2g-D (25%), a substantial amount of 2g-D′ (75%) was obtained. In contrast, no aromatic deuteriation was observed when phenanthrene (6) was used as the substrate under the same conditions. The formation of 2g-D′ could be attributed to the isomerization of benzylic radical species (also see the ESI, Fig. S1^†). This preliminary evidence pointed at a radical mechanism, although a carbocation intermediate cannot be ruled out completely.Open in a separate windowScheme 8Control experiments.Consistent with literature reports on BR₃,^43,44 the reactivity of BBr₃ towards homolytic debromination decreases sharply along the series BBr₃, BBr₂OR, and BBr(OR)₂ as a consequence of π-bonding between oxygen and boron. With 0.5 equiv BBr₃, only 21% of 2a was obtained even with a prolonged reaction time of 24 h. These data indicated that only the first equivalent of Br from BBr₃ is crucial for the reactivity, and contribution from the possible BBr_a(OR)_3−a byproducts should be insignificant.A series of ¹H and ¹¹B NMR experiments were conducted to gain further insight. Upon mixing BBr₃ with 1a in the absence of O₂ and a proton source, both 1a and BBr₃ were mostly consumed, and a new ¹¹B signal at 64 ppm (see the ESI, Fig. S2^†) emerged as a singlet, which is characteristic of an alkyldihaloborane species.^45,46 From ¹H NMR, it is clear that 1a is ring-opened (see the ESI, Fig. S4^†), and the species has a similar NMR pattern to a hydroborated cyclopropane, which has been reported as a reaction intermediate in literature examples (see the ESI, Fig. S3^†).³⁷ Direct bromoboration of alkynes or allenes with BBr₃ is well documented.^47,48 While this new species cannot be clearly identified, it is speculated that it could be the direct bromoboration product or hydroxyboration product. Nevertheless, it is clear that the interaction between BBr₃ and 1a does not lead to the anti-Markovnikov product 2a in the absence of O₂ and a proton source.When i-PrOH and BBr₃ were mixed in CD₂Cl₂ under air, the ¹¹B signal of BBr₃ (39 ppm) disappeared and a new signal at 25.0 ppm emerged. A new proton signal at −2.68 ppm also appeared in the ¹H NMR study of the same sample. The two new signals (25.0 ppm in ¹¹B NMR and −2.68 ppm in ¹H NMR) diminished gradually upon the addition of 1a and the amount of anti-Markovnikov product 2a increased accordingly (see the ESI, Fig. S5^†). On the other hand, a new ¹¹B NMR signal at 18.9 ppm (but no signal at 25.0 ppm) was observed when the same mixture was prepared in the absence of O₂ and attributed to the formation of the Lewis adduct between i-PrOH and BBr₃ (see the ESI, Fig. S6^†). Thus, it is reasonable to propose that the active species, responsible for initiating the anti-Markovnikov hydrobromination of cyclopropanes, was formed only in the presence of O₂.A DFT computational study was also performed to shed light on the mechanism (Fig. 1). While there are no reports on radical reactions triggered by BBr₃/O₂, we speculate that the reaction mechanism might be analogous to the classical BR₃/O₂ system in which the putative peroxy-boron species A is generated⁴⁹ at the initiation stage of the radical process (Fig. 1A) and corresponds to the new NMR signals (25.0 ppm in ¹¹B NMR and −2.68 ppm in ¹H NMR).^3,50,51 Based on the calculated energy profile, species A is capable of brominating cyclopropane 1a through a radical mechanism to give B (Fig. 1B). It is also calculated that A and A′ could be in equilibrium, but species A (ΔG = −7.6 kcal mol⁻¹) was found to be a more competent Br donor than A′ (ΔG = 0.6 kcal mol⁻¹) in the halogen atom transfer (XAT), potentially due to the intramolecular hydrogen bond that stabilizes the by-product I (Fig. 1C). It was also calculated that BBr₃ can reversibly react with water to give adduct C (¹¹B NMR signal = 18.9 ppm). Species C is unable to serve as a Br radical donor to brominate cyclopropane 1a (ΔG = 69.6 kcal mol⁻¹).Open in a separate windowFig. 1Reaction mechanism. (A) Plausible reaction pathways. (B) Calculated free energy profile of the anti-Markovnikov hydrobromination of 1a at the ωB97X-D/6-311++G(d,p), SMD(CH₂Cl₂)//ωB97X-D/6-31+G(d,p) level of theory. (C) Potential competing pathways.However, species C is capable of acting as a hydrogen radical donor to species B, furnishing the desired product 2a. This result is in alignment with the proposal in the literature in which trialkylborane-ROH complexes (R = H, Me) might act as H-donors as a result of the weakened O–H bond.⁵² Species D, which is formed from species C after the hydrogen atom transfer (HAT), was calculated to be a competent Br radical donor to brominate cyclopropane 1a to give B, thereby propagating the radical chain. Thus, we propose that oxygen is required only in the initiation stage for the generation of species A, while species C and D are responsible for propagation. Indeed, the reaction was sluggish under an inert atmosphere, while the re-introduction of oxygen into a system initially free of oxygen triggered the anti-Markovnikov hydrobromination (see the ESI, Scheme S2^†). The HAT from species A to 1a was also explored computationally, but species A could not be optimized as a stable energy minimum. Species C may also serve as a hydrogen radical donor and react with cyclopropane 1a to give species D and E, which would go on to produce the Markovnikov product 3a. However, this hydrogen atom transfer reaction is endergonic by 31.2 kcal mol⁻¹ (Fig. 1C), making it a minor pathway compared to the competing hydrogen atom transfer from C to B that gives D and 2a (Fig. 1A). This result is consistent with the experimental observation that the Markovnikov product 3a became dominant when the reaction was conducted under an inert atmosphere (Scheme 8A).In the ¹H NMR study of the reaction using 1a, apart from 2a, 3a and 4a (Scheme 2), a trace amount of allylbenzene was detected initially and diminished over time. We speculate that the allylbenzene (Fig. 1A, species G) might be formed through the zwitterionic species F as proposed in the recent studies by Wang and Shi.^37,53 The eventual disappearance of allylbenzene could be attributed to the radical bromination to give species H and the subsequent formation of 2a.In a deuterium labeling experiment with 1a as the substrate and D₂O as the deuterium source, we observed exclusive deuterium incorporation at the benzylic carbon to give product 2a-D, potentially through the C(1) radical species B1 (Scheme 9, eqn (1)). However, the deuterium incorporation pattern is vastly different when using allylbenzene instead of 1a, for which C(2) deuterated product 2a-D′ was obtained predominately (Scheme 9, eqn (2)) (also see the ESI, Fig. S7^†). The formation of 2a-D′ from allylbenzene may proceed through the C(2) radical species H1. A small amount of 2a-D (9%) was also detected in the reaction with allylbenzene, attributed to the slow 1,2-hydrogen shift⁵⁴ converting H1 to the more stable benzylic radical B1. These results suggest that the 1,2-hydrogen shift between the radical species H and B (Fig. 1A) should be much slower than the radical protonation process, implying that allylbenzene is unlikely to be the key intermediate in the reaction.Open in a separate windowScheme 9Mechanistic insights from deuteriobromination.     

Ferro-self-assembly: magnetic and electrochemical adaptation of a multiresponsive zwitterionic metalloamphiphile showing a shape-hysteresis effect

Metallosurfactants are molecular compounds which combine the unique features of amphiphiles, like their capability of self-organization, with the peculiar properties of metal complexes like magnetism and a rich redox chemistry. Considering the high relevance of surfactants in industry and science, amphiphiles that change their properties on applying an external trigger are highly desirable. A special feature of the surfactant reported here, 1-(Z)-heptenyl-1′-dimethylammonium-methyl-(3-sulfopropyl)ferrocene (6), is that the redox-active ferrocene constituent is in a gemini-position. Oxidation to 6⁺ induces a drastic change of the surfactant''s properties accompanied by the emergence of paramagnetism. The effects of an external magnetic field on vesicles formed by 6⁺ and the associated dynamics were monitored in situ using a custom-made optical birefringence and dual dynamic light scattering setup. This allowed us to observe the optical anisotropy as well as the anisotropy of the diffusion coefficient and revealed the field-induced formation of oriented string-of-pearls-like aggregates and their delayed disappearance after the field is switched off.

The self-organization properties of a stimuli responsive amphiphile can be altered by subjecting the paramagnetic oxidized form to a magnetic field of 0.8 T and monitored in real time by coupling optical birefringence with dynamic light scattering.

Amphiphiles (or surfactants) combine hydrophilic (the so-called headgroups) and lipophilic entities (the so-called tails) as integral parts of their molecular structures. This particular construction principle provides them with the ability to display concentration-dependent self-organization in nonpolar and polar solvents.¹ Amphiphiles with advanced functions that go far beyond the traditional ones as emulsifiers, stabilizing agents for interfaces, or detergents were meanwhile realized by skillful manipulation of any of its constituents.^2–4 Recent examples are micellar LEDs,^5,6 catalysts,^7–9 or batteries.¹⁰ Such applications are important hallmarks on the way to even more sophisticated amphiphiles such as the ones found in nature, e.g. in the pockets of enzymes.^11–18 An important milestone is the advent of (multi-) stimuli-responsive amphiphiles, whose encoded functionalities respond to (different) external triggers. Such systems are capable of adaptive self-assembly, which can be controlled using an external input such as the pH, temperature, ionic strength, or redox state.^19–26Paramagnetic amphiphiles, recently reviewed by Eastoe and coworkers, constitute a fascinating family of stimuli-responsive surfactants.²⁷ Particular attention has been paid to magnetic ionic liquids based on amphiphilic transition metal complexes, as their properties are often superior to those of conventional magnetic fluids (ferrofluids).^28–31 Self-assembly results in high effective concentrations of the paramagnetic metal centers, and this in turn allows us to control their physico-chemical properties and the morphologies of their superstructures through an external magnetic field. Such a scheme has the added advantage that the external stimulus is non-invasive. In many current realizations of such systems, however, the magneto-active (transition) metal ion is only present as a constituent of the counterion of a cationic surfactant, but is not an integral constituent of the surfactant itself.^21,30,31Some of us have previously reported redox-switchable as well as paramagnetic stimuli-responsive amphiphiles of relevance to the current work.^32,33 We thought that ferrocene would be an ideal building block in order to combine both these kinds of stimuli within one single amphiphile.^34–37 On oxidation, the diamagnetic, hydrophobic ferrocene nucleus is transformed into a paramagnetic S = 1/2 ferrocenium ion with a distinct hydrophilic character.^38–41 Oxidation does hence not only generate a magnetic moment, but also transfers the ferrocene nucleus from the lipo- to the hydrophilic part of the amphiphile, thereby changing its entire structure. A 1,1′-disubstitution pattern of the ferrocene scaffold, which is synthetically well accessible,^34,42–44 seemed particularly suited for such an endeavor.Studies on paramagnetic amphiphiles are often thwarted by the non-trivial analytics involved in their characterization. Detailed investigations often rely on small-angle neutron scattering (SANS), which is time-consuming and costly and suffers from poor availability.^{27,30,31,45–47} Moreover, SANS is only of limited value for following kinetically fast processes which would be desirable for the live monitoring of structural changes occurring in solution. Optical birefringence is a well-established method to monitor the dynamic response of materials to external fields.^48–50 Although of high intrinsic value, optical birefringence measurements in magnetic fields were only rarely applied for the study of paramagnetic amphiphiles.²⁹We here report the zwitterionic, ferrocene-based amphiphile FcNMe₂SO₃Heptene 6 (see Fig. 1, Fc = ferrocenyl) with a sultone headgroup. Compound 6 is unique in that its self-assembly properties can be controlled by three different external stimuli, namely the (i) addition of an electrolyte, (ii) addition of an oxidant/reductant, and (iii) exposure to an external magnetic field. We also demonstrate that optical birefringence in combination with dynamic light scattering (DLS) measurements in two orthogonal directions provides detailed insights into the functional response of aggregated magnetic nanoparticles formed by 6⁺ to an external magnetic field in real time. Specifically, we have observed the formation of string-of-pearls-like aggregates of 6⁺ in a magnetic field (0.8 T), the field-induced anisotropy of the diffusion of aggregated nanoparticles, and a hysteresis effect for their disappearance after the magnetic field is switched off. Thus, the anisotropy of larger aggregates persists for more than 5 min, while the structural alignment of smaller ones vanishes at a significantly faster rate.Open in a separate windowFig. 1Synthesis of FcNMe₂SO₃Heptene (6). (a) Synthesis of 6; (b) molecular structure of 6 crystallized from acetonitrile. C; dark grey, N; turquoise, Fe; orange, S; yellow, O; red, H atoms are omitted for clarity.     

A Paal–Knorr agent for chemoproteomic profiling of targets of isoketals in cells

Natural systems produce various γ-dicarbonyl-bearing compounds that can covalently modify lysine in protein targets via the classic Paal–Knorr reaction. Among them is a unique class of lipid-derived electrophiles – isoketals that exhibit high chemical reactivity and critical biological functions. However, their target selectivity and profiles in complex proteomes remain unknown. Here we report a Paal–Knorr agent, 4-oxonon-8-ynal (herein termed ONAyne), for surveying the reactivity and selectivity of the γ-dicarbonyl warhead in biological systems. Using an unbiased open-search strategy, we demonstrated the lysine specificity of ONAyne on a proteome-wide scale and characterized six probe-derived modifications, including the initial pyrrole adduct and its oxidative products (i.e., lactam and hydroxylactam adducts), an enlactam adduct from dehydration of hydroxylactam, and two chemotypes formed in the presence of endogenous formaldehyde (i.e., fulvene and aldehyde adducts). Furthermore, combined with quantitative chemoproteomics in a competitive format, ONAyne permitted global, in situ, and site-specific profiling of targeted lysine residues of two specific isomers of isoketals, levuglandin (LG) D2 and E2. The functional analyses reveal that LG-derived adduction drives inhibition of malate dehydrogenase MDH2 and exhibits a crosstalk with two epigenetic marks on histone H2B in macrophages. Our approach should be broadly useful for target profiling of bioactive γ-dicarbonyls in diverse biological contexts.

Natural systems produce various γ-dicarbonyl-bearing compounds that can covalently modify lysine in protein targets via the classic Paal–Knorr reaction.

Synthetic chemistry methods have been increasingly underscored by their potential to be repurposed as biocompatible methods for both chemical biology and drug discovery. The most-known examples of such a repurposing approach include the Staudinger ligation¹ and the Huisgen-based click chemistry.² Moreover, bioconjugation of cysteine and lysine can be built upon facile chemical processes,³ while chemoselective labelling of other polar residues (e.g., histidine,⁴ methionine,⁵ tyrosine,⁶ aspartic and glutamic acids^7,8) requires more elaborate chemistry, thereby offering a powerful means to study the structure and function of proteins, even at a proteome-wide scale.The classical Paal–Knorr reaction has been reported for a single-step pyrrole synthesis in 1884.^9,10 The reaction involves the condensation of γ-dicarbonyl with a primary amine under mild conditions (e.g., room temperature, mild acid) to give pyrrole through the intermediary hemiaminals followed by rapid dehydration of highly unstable pyrrolidine adducts (Fig. S1^†).Interestingly, we and others have recently demonstrated that the Paal–Knorr reaction can also readily take place in native biological systems.^11–13 More importantly, the Paal–Knorr precursor γ-dicarbonyl resides on many endogenous metabolites and bioactive natural products.¹⁴ Among them of particular interest are isoketals¹⁵ (IsoKs, also known as γ-ketoaldehydes) which are a unique class of lipid derived electrophiles (LDEs) formed from lipid peroxidation (Fig. S2^†)¹⁶ that has emerged as an important mechanism for cells to regulate redox signalling and inflammatory responses,¹⁷ and drive ferroptosis,¹⁸ and this field has exponentially grown over the past few years. It has been well documented that the γ-dicarbonyl group of IsoKs can rapidly and predominantly react with lysine via the Paal–Knorr reaction to form a pyrrole adduct in vitro (Fig. 1).¹⁵ Further, the pyrrole formed by IsoKs can be easily oxidized to yield lactam and hydroxylactam products in the presence of molecular oxygen (Fig. 1). These rapid reactions are essentially irreversible. Hence, IsoKs react with protein approximately two orders of magnitude faster than the most-studied LDE 4-hydoxynonenal (4-HNE) that contains α,β-unsaturated carbonyl to generally adduct protein cysteines by Michael addition (Fig. S3^†).¹⁵ Due to this unique adduction chemistry and rapid reactivity, IsoKs exhibit intriguing biological activities, including inhibition of the nucleosome complex formation,¹⁹ high-density lipoprotein function,²⁰ mitochondrial respiration and calcium homeostasis,²¹ as well as activation of hepatic stellate cells.²² Furthermore, increases in IsoK-protein adducts have been identified in many major diseases,²³ such as atherosclerosis, Alzheimer''s disease, hypertension and so on.Open in a separate windowFig. 1The Paal–Knorr precursor γ-dicarbonyl reacts with the lysine residue on proteins to form diverse chemotypes via two pathways. The red arrow shows the oxidation pathway, while the blue one shows the formaldehyde pathway.Despite the chemical uniqueness, biological significance, and pathophysiological relevance of IsoKs, their residue selectivity and target profiles in complex proteomes remain unknown, hampering the studies of their mechanisms of action (MoAs). Pioneered by the Cravatt group, the competitive ABPP (activity-based protein profiling) has been the method of choice to analyse the molecular interactions between electrophiles (e.g., LDEs,²⁴ oncometabolites,²⁵ natural products,^26,27 covalent ligands and drugs^28–30) and nucleophilic amino acids across complex proteomes. In this regard, many residue-specific chemistry methods and probes have been developed for such studies. For example, several lysine-specific probes based on the activated ester warheads (e.g., sulfotetrafluorophenyl, STP;³¹N-hydroxysuccinimide, NHS³²) have recently been developed to analyse electrophile–lysine interactions at a proteome-wide scale in human tumour cells, which provides rich resources of ligandable sites for covalent probes and potential therapeutics. Although these approaches can also be presumably leveraged to globally and site-specifically profile lysine-specific targets IsoKs, the reaction kinetics and target preference of activated ester-based probes likely differ from those of γ-dicarbonyls, possibly resulting in misinterpretation of ABPP competition results. Ideally, a lysine profiling probe used for a competitive ABPP analysis of IsoKs should therefore possess the same, or at least a similar, warhead moiety. Furthermore, due to the lack of reactive carbonyl groups on IsoK-derived protein adducts, several recently developed carbonyl-directed ligation probes for studying LDE-adductions are also not suitable for target profiling of IsoKs.^33,34Towards this end, we sought to design a “clickable” γ-dicarbonyl probe for profiling lysine residues and, in combination with the competitive ABPP strategy, for analysing IsoK adductions in native proteomes. Considering that the diversity of various regio- and stereo- IsoK isomers¹⁵ (a total of 64, Fig. S2^†) in chemical reactivity and bioactivities is likely attributed to the substitution of γ-dicarbonyls at positions 2 and 3, the “clickable” alkyne handle needs to be rationally implemented onto the 4-methyl group in order to minimize the biases when competing with IsoKs in target engagement. Interestingly, we reasoned that 4-oxonon-8-ynal, a previously reported Paal–Knorr agent used as an intermediate for synthesizing fatty acid probes³⁵ or oxa-tricyclic compounds,³⁶ could be repurposed for the γ-dicarbonyl-directed ABPP application. With this chemical in hand (herein termed ONAyne, Fig. 2A), we first used western blotting to detect its utility in labelling proteins, allowing visualization of a dose-dependent labelling of the proteome in situ (Fig. S4^†). Next, we set up to incorporate this probe into a well-established chemoproteomic workflow for site-specific lysine profiling in situ (Fig. 2A). Specifically, intact cells were labelled with ONAyne in situ (200 μM, 2 h, 37 °C, a condition showing little cytotoxicity, Fig. S5^†), and the probe-labelled proteome was harvested and processed into tryptic peptides. The resulting probe-labelled peptides were conjugated with both light and heavy azido-UV-cleavable-biotin reagents (1 : 1) via Cu^I-catalyzed azide–alkyne cycloaddition reaction (CuAAC, also known as click chemistry). The biotinylated peptides were enriched with streptavidin beads and photoreleased for LC-MS/MS-based proteomics. The ONAyne-labelled peptides covalently conjugated with light and heavy tags would yield an isotopic signature. We considered only those modified peptide assignments whose MS1 data reflected a light/heavy ratio close to 1.0, thereby increasing the accuracy of these peptide identifications. Using this criterium, we applied a targeted database search to profile three expected probe-derived modifications (PDMs), including 13 pyrrole peptide adducts (Δ273.15), 77 lactam peptide adducts (Δ289.14), and 557 hydroxylactam peptide adducts (Δ305.14), comprising 585 lysine residues on 299 proteins (Fig. S6 and S7^†). Among them, the hydroxylactam adducts were present predominately, since the pyrrole formed by this probe, the same as IsoKs, can be easily oxidized when being exposed to O₂. This finding was in accordance with a previous report where the pyrrole adducts formed by the reaction between IsoK and free lysine could not be detected, but rather their oxidized forms.³⁷ Regardless, all three types of adducts were found in one lysine site of EF1A1 (K387, Fig. S8^†), further confirming the intrinsic relationship among those adductions in situ.Open in a separate windowFig. 2Adduct profile and proteome-wide selectivity of the γ-dicarbonyl probe ONAyne. (A) Chemical structure of ONAyne and schematic workflow for identifying ONAyne-adducted sites across the proteome. (B) Bar chart showing the distribution of six types of ONAyne-derived modifications formed in situ and in vitro (note: before probe labelling, small molecules in cell lysates were filtered out through desalting columns).State-of-the-art blind search can offer an opportunity to explore unexpected chemotypes (i.e., modifications) derived from a chemical probe and to unbiasedly assess its proteome-wide residue selectivity.^38,39 We therefore sought to use one of such tools termed pChem³⁸ to re-analyse the MS data (see Methods, ESI^†). Surprisingly, the pChem search identified three new and abundant PDMs (Fig. 1 and Table S1^†), which dramatically expand the ONAyne-profiled lysinome (2305 sites versus 585 sites). Overall, these newly identified PDMs accounted for 74.6% of all identifications (Fig. 2B and Table S2^†). Among them, the PDM of Δ287.13 (Fig. 1 and S7^†) might be an enlactam product via dehydration of the probe-derived hydroxylactam adduct. The other two might be explained by the plausible mechanism as follows (Fig. 1). The endogenous formaldehyde (FA, produced in substantial quantities in biological systems) reacts with the probe-derived pyrrole adduct via nucleophilic addition to form a carbinol intermediate, followed by rapid dehydration to a fulvene (Δ285.15, Fig. S7^†) and immediate oxidation to an aldehyde (Δ301.14, Fig. S7^†). In line with this mechanism, the amount of FA-derived PDMs was largely eliminated when the in vitro ONAyne labelling was performed in the FA-less cell lysates (Fig. 2B and Table S3^†). Undoubtedly, the detailed mechanisms underlying the formation of these unexpected PDMs require further investigation, and so does the reaction kinetics. Regardless, all main PDMs from ONAyne predominantly target the lysine residue with an average localization probability of 0.77, demonstrating their proteome-wide selectivity (Fig. S9^†).Next, we adapted an ABPP approach to globally and site-specifically quantify the reactivity of lysine towards the γ-dicarbonyl warhead through a dose-dependent labelling strategy (Fig. 3A) that has been proved to be successful for other lysine-specific probes (e.g., STP alkyne).³¹ Specifically, MDA-MB-231 cell lysates were treated with low versus high concentrations of ONAyne (1 mM versus 0.1 mM) for 1 h. Probe-labelled proteomes were digested into tryptic peptides that were then conjugated to isotopically labelled biotin tags via CuAAC for enrichment, identification and quantification. In principle, hyperreactive lysine would saturate labelling at the low probe concentration, whereas less reactive ones would show concentration-dependent increases in labelling. For fair comparison, the STP alkyne-based lysine profiling data were generated by using the same chemoproteomic workflow. Although 77.5% (3207) ONAyne-adducted lysine sites can also be profiled by STP alkyne-based analysis, the former indeed has its distinct target-profile with 930 lysine sites newly identified (Fig. S10 and Table S4^†). Interestingly, sequence motif analysis with pLogo⁴⁰ revealed a significant difference in consensus motifs between ONAyne- and STP alkyne-targeting lysines (Fig. S11^†).Open in a separate windowFig. 3ONAyne-based quantitative reactivity profiling of proteomic lysines. (A) Schematic workflow for quantitative profiling of ONAyne–lysine reactions using the dose-dependent ABPP strategy (B) Box plots showing the distribution of R_10:1 values quantified in ONAyne- and STP alkyne-based ABPP analyses, respectively. Red lines showing the median values. ***p ≤ 0.001 two-tailed Student''s t-test. (C) Representative extracted ion chromatograms (XICs) showing changes in the EF1A1 peptide bearing K273 that is adducted as indicated, with the profiles for light and heavy-labelled peptides in blue and red, respectively.Moreover, we quantified the ratio (R_{1 mM:0.1 mM}) for a total of 2439 ONAyne-tagged lysines (on 922 proteins) and 17904 STP alkyne-tagged lysines (on 4447 proteins) across three biological replicates (Fig. S12 and Table S5^†). Strikingly, only 26.7% (651) of quantified sites exhibited nearly dose-dependent increases (R_{1 mM:0.1 mM} > 5.0) in reactivity with ONAyne, an indicative of dose saturation (Fig. 3B and C). In contrast, such dose-dependent labelling events accounted for >69.1% of all quantified lysine sites in the STP alkyne-based ABPP analysis.³¹ This finding is in accordance with the extremely fast kinetics of reaction between lysine and γ-dicarbonyls (prone to saturation). Nonetheless, by applying 10-fold lower probe concentrations, overall 1628 (80.2%) detected lysines could be labelled in a fully concentration-dependent manner with the median R_10:1 value of 8.1 (Fig. 3B, C, S12 and Table S5^†). Next, we asked whether the dose-depending quantitation data (100 μM versus 10 μM) can be harnessed to predict functionality. By retrieving the functional information for all quantified lysines from the UniProt Knowledgebase, we found that those hyper-reactive lysines could not be significantly over-represented with annotation (Fig. S12^†). Nonetheless, among all quantified lysines, 509 (25.1%) possess functional annotations, while merely 2.5% of the human lysinome can be annotated. Moreover, 381 (74.8%) ONAyne-labelled sites are known targets of various enzymatic post-translational modifications (PTMs), such as acetylation, succinylation, methylation and so on (Fig. S13^†). In contrast, all known PTM sites accounted for only 59.6% of the annotated human lysinome. These findings therefore highlight the intrinsic reactivity of ONAyne towards the ‘hot spots’ of endogenous lysine PTMs.The aforementioned results validate ONAyne as a fit-for-purpose lysine-specific chemoproteomic probe for competitive isoTOP-ABPP application of γ-dicarbonyl target profiling. Inspired by this, we next applied ONAyne-based chemoproteomics in an in situ competitive format (Fig. 4A) to globally profile lysine sites targeted by a mixture of levuglandin (LG) D₂ and E₂, two specific isomers of IsoKs that can be synthesized conveniently from prostaglandin H₂ (ref. 41) (Fig. S2^†). Specifically, mouse macrophage RAW264.7 cells (a well-established model cell line to study LDE-induced inflammatory effects) were treated with 2 μM LGs or vehicle (DMSO) for 2 h, followed by ONAyne labelling for an additional 2 h. The probe-labelled proteomes were processed as mentioned above. For each lysine detected in this analysis, we calculated a control/treatment ratio (R_C/T). Adduction of a lysine site by LGs would reduce its accessibility to the ONAyne probe, and thus a higher R_C/T indicates increased adduction. In total, we quantified 2000 lysine sites on 834 proteins across five biological replicates. Among them, 102 (5.1%) sites exhibited decreases of reactivity towards LGs treatment (P < 0.05, Table S6^†), thereby being considered as potential targets of LGs. Notably, we found that different lysines on the same proteins showed varying sensitivity towards LGs (e.g., LGs targeted K3 of thioredoxin but not K8, K85 and K94, Table S6^†), an indicative of changes in reactivity, though we could not formally exclude the effects of changes in protein expression on the quantified competition ratios. Regardless, to the best of our knowledge, the proteome-wide identification of potential protein targets by IsoKs/LGs has not been possible until this work.Open in a separate windowFig. 4ONAyne-based in situ competitive ABPP uncovers functional targets of LGs in macrophages. (A) Schematic workflow for profiling LGs–lysine interactions using ONAyne-based in situ competitive ABPP. (B) Volcano plot showing the log₂ values of the ratio between the control (heavy) and LGs-treated (light) channels and the −log₁₀(P) of the statistical significance in a two-sample t-test for all quantified lysines. Potential targets of LGs are shown in blue (R_C/T>1.2, P < 0.05), with the validated ones in red. (C) Bar chart showing the inhibitory effect of 2 μM LGs on the cellular enzymatic activity of MDH2. Data represent means ± standard deviation (n = 3). Statistical significance was calculated with two-tailed Student''s t-tests. (D) Pretreatment of LGs dose-dependently blocked ONAyne-labelling of MDH2 in RAW264.7 cells, as measured by western blotting-based ABPP. (E and F) LGs dose-dependently decreased the H2BK5 acetylation level in RAW 264.7 cells, as measured either by western blotting (E) or by immunofluorescence imaging (F). n = 3. For G, nuclei were visualized using DAPI (blue).We initially evaluated MDH2 (malate dehydrogenase, mitochondrial, also known as MDHM), an important metabolic enzyme that possesses four previously uncharacterized liganded lysine sites (K157, K239, K301 and K329, Fig. 4B) that are far from the active site (Fig. S14^†). We found that LGs dramatically reduced the catalytic activity of MDH2 in RAW264.7 cells (Fig. 4C), suggesting a potentially allosteric effect. We next turned our attention to the targeted sites residing on histone proteins, which happen to be modified by functionally important acetylation, including H2BK5ac (Fig. 4B) that can regulate both stemness and epithelial–mesenchymal transition of trophoblast stem cells.⁴² We therefore hypothesized that rapid adduction by LGs competes with the enzymatic formation of this epigenetic mark. Immunoblotting-based competitive ABPP confirmed that LGs dose-dependently blocked probe labelling of H2B (Fig. 4D). Further, both western blots and immunofluorescence assays revealed that LG treatment decreased the level of acetylation of H2BK5 (average R_C/T = 1.3, P = 0.007) in a concentration-dependent manner (Fig. 4E and F). Likewise, a similar competitive crosstalk was observed between acetylation and LG-adduction on H2BK20 (average R_C/T = 1.2, P = 0.01) that is required for chromatin assembly⁴³ and/or gene regulation⁴⁴ (Fig. 4B and S15^†). Notably, these findings, together with several previous reports by us and others about histone lysine ketoamide adduction by another important LDE, 4-oxo-2-noenal,^11,45,46 highlight again the potentially important link between lipid peroxidation and epigenetic regulation. In addition to the targets validated as above, many other leads also merit functional studies considering diverse biological or physiologic effects of LGs in macrophages.     

CsAlB3O6F: a beryllium-free deep-ultraviolet nonlinear optical material with enhanced thermal stability

The design of new beryllium-free deep-ultraviolet nonlinear optical materials is important but challenging. Here, we describe a new strategy to search for such materials based on rational selection of fundamental structural units. By combining asymmetric AlO₃F tetrahedra and π-conjugated B₃O₆ rings, a new aluminum borate fluoride, CsAlB₃O₆F was obtained. It exhibits excellent linear and nonlinear optical properties including a high optical transmittance with a cut-off edge shorter than 190 nm, large second harmonic generation intensities (2.0× KH₂PO₄, KDP), and suitable birefringence for phase-matching under 200 nm. It also has good thermal stability and can be synthesized easily in an open system.

A new potential deep-ultraviolet nonlinear optical material CsAlB₃O₆F was designed by a rational selection of fundamental structural units. This material does not require toxic raw materials and can be grown in an open system.

The exploration of new deep-ultraviolet (DUV) nonlinear optical (NLO) materials is intriguing and of great importance because these materials are crucial for the development of all-solid-state DUV lasers.^1–3 For NLO materials, the main obstacles to DUV application are 3-fold:^4,5 (i) a wide DUV transparency window (wavelength cut-off edge < 200 nm), (ii) high NLO coefficients (>1× commercial KH₂PO₄, KDP), and (iii) sufficient birefringence to satisfy phase matching conditions in the DUV region. Until now, only KBe₂BO₃F₂ (KBBF) can certainly break through these barriers and generate lasers with wavelengths shorter than 200 nm by direct second harmonic generation (SHG).⁶ However, KBBF is limited in practical use because of its adverse layered crystal growth habit and use of the highly toxic beryllium component. To find a KBBF replacement, many new NLO crystals have been developed continuously, but have so far been unable to achieve desired NLO properties.^7–10Recently, it was shown that fluorooxoborates and fluorophosphates with mixed O/F anionic groups might be recognized as new sources for discovering DUV NLO materials. For example, Pan''s group proposed that the [BO_xF_4−x] (x = 1, 2, 3) tetrahedra are good units to balance the multiple criteria of DUV NLO materials.¹¹ Accordingly, monofluorophosphates with non-π-conjugated asymmetric [PO₃F] units were also paid attention, exhibiting superior optical properties.^12,13 Thereafter, numerous fluorooxoborates and fluorophosphates with an unprecedented crystal structure and high performance as potential DUV NLO materials have been reported.^14–17 Nevertheless, because these materials are relatively unstable at high temperature in air, one need to develop a suitable crystal growth method under sealed conditions and/or search for a suitable flux (solvent) system at relatively low temperature.^14,16 Alternatively, it is possible to achieve a beryllium-free, higher stability DUV NLO crystal by combining other anionic groups, such as [AlO₄], [PO₄], [SiO₄], and [ZnO₄].^18–21Enlightened by the successful synthesis of NLO-active fluorooxoborates and fluorophosphates, we proposed that aluminum borate fluorides with [AlO_mF_n] (m + n = 4, 5, 6, AlOF for short) units might be another choice for exploring new NLO materials. Four specific aspects were considered: (I) aluminum borates are preferred because of the chemical and coordination environment similarity between Al³⁺ and Be²⁺ cations.²² (II) similar to fluorooxoborates, the substitution of O²⁻ with larger electronegativity F⁻ can increase the bandgap and optical anisotropy due to the ionicity of the Al–F bond. (III) The rich coordination environment of AlOF groups provides more structural possibilities than those of other mixed-anionic groups (e.g. [BeO₃F], [BO₃F], [BO₂F₂], see Fig. S1^†). Different from the B or Be atom, the Al atom has empty d orbitals, and it can form sp³, sp³d, and sp³d² hybrid orbitals when bonding with O/F atoms. Consequently, diverse AlOF groups without anion-site disorder, such as [AlO₃F] tetrahedra, [AlO₃F₂] or [AlO₄F] trigonal bipyramids, and [AlO₅F], [AlO₄F₂], [AlO₂F₄], or [AlOF₅] octahedra, have been achieved (Fig. S1^†). Replacing oxygen atoms with fluorine in the Al–O polyhedra not only increases the degree of freedom (e.g. cis/trans conformation), but also causes a (local) symmetry breaking and results in increased microscopic susceptibility and optical anisotropy, which are beneficial to build a noncentrosymmetric (NCS) material. (IV) Aluminum borate fluorides without B–F bonds (e.g. BaAlBO₃F₂,²³ Rb₃Al₃B₃O₁₀F,²⁴ and K₃Ba₃Li₂Al₄B₆O₂₀F²⁰) show good thermal stability, and large crystals could be obtained in air.By applying the strategy described above, we tried to design a new DUV NLO material according to the blueprints shown in Fig. 1a. From KBBF to fluorooxoborates (e.g. NH₄B₄O₆F,²⁵ ABF), the nontoxic [BO₃F] units were selected to replace the [BeO₃F] tetrahedra of KBBF while the NLO properties were retained. Meanwhile, benzene-like [B₃O₆] rings (the same as that in β-BaB₂O₄, BBO) with a better conjugated π-orbital system were utilized to replace [BO₃] triangles (the case in CsB₄O₆F,²⁶ CBF), which could contribute to the increase of SHG responses compared to KBBF. In this work, more stable [AlO₃F] units combined with the [B₃O₆] rings were chosen as fundamental building units (FBUs) and a new potential DUV NLO material CsAlB₃O₆F (CABF) has been successfully synthesized. The material can be obtained easily in an open system and it exhibits a short UV cutoff edge below 190 nm with a powder SHG response of 2.0× KDP under 1064 nm incident radiation. The first-principles calculations reveal that CABF possesses moderate SHG coefficients and sufficient birefringence for DUV phase-matching. With the introduction of new AlO₃F units, CABF continues to maintain the excellent structural features of both KBBF and β-BBO, and thus exhibits a great potential to be a Be-free DUV material for nonlinear light–matter interactions.Open in a separate windowFig. 1(a) The structural evolution from KBe₂BO₃F₂ to CsAlB₃O₆F. (b) The 2D [AlB₃O₆F]_∞ layer of CABF. (c) Crystal structure of CABF.Single crystals of CABF were obtained by the conventional flux method in an open system (see the Experimental details in the ESI^†). CABF crystallizes in the NCS orthorhombic polar space group Pna2₁ (Tables S1–S3^†). Its crystal structure is built up from [AlB₃O₆F]_∞ 2D layers and the Cs⁺ cations reside in the interlayer space (Fig. 1b and c). The FBUs of CABF are the [AlB₃O₈F] groups composed of one [B₃O₆] ring and one [AlO₃F] tetrahedron. Three crystallographically independent boron atoms are all three-coordinated to oxygen atoms forming the BO₃ triangles, which further form a B₃O₆ ring by sharing their terminal O atoms. The B–O bond distances and the O–B–O angles are in the range of 1.342(10) to 1.395(11) Å and 117.2(7) to 121.9(8)°, respectively. The Al³⁺ cation is coordinated to three oxygen atoms (Al–O bond lengths: 1.719(6)–1.734(6) Å) and one fluorine atom (Al–F bond length equals 1.675(5) Å) to form a distorted AlO₃F tetrahedron. In the bc-plane, the [AlO₃F] tetrahedra bond with discrete [B₃O₆] rings to create [AlB₃O₆F]_∞ layers that contain 18-membered rings (Fig. 1b). In one layer, the apical F atoms of [AlO₃F] tetrahedra point upward and downward regularly. The Cs⁺ cations are twelve-coordinated, forming the [CsO₁₁F] polyhedra with a Cs–F distance of 3.292(7) Å and Cs–O distances ranging from 3.256(5) to 3.585(6) Å (Fig. S2^†). Because the Cs⁺ cations reside in the tunnels created by the 18-membered rings and adjacent [AlB₃O₆F]_∞ layers, the interlayer distance in CABF is 4.03 Å. Compared to the distance of two adjacent layers in KBBF (6.25 Å), CABF should exhibit stronger interlayer interactions. Theoretical calculation based on the density functional theory (DFT) also confirms that CABF shows a larger interlayer binding energy than KBBF (0.178 vs. 0.0314 eV Å⁻² per layer). These results indicate that CABF might possess a better crystal growth habit without layering. The bond valence sum calculation²⁷ (Table S2^†) and IR spectrum (Fig. S3^†) indicate that all atoms of CABF have the expected oxidation states and coordination environments.Polycrystalline CABF was synthesized by a stoichiometric solid-state reaction of CsF, Al₂O₃ and H₃BO₃ at 500 °C, and the phase purity was confirmed by powder X-ray diffraction (PXRD) Rietveld refinement (Fig. S4^†). To check the stability of CABF, thermal measurements and further PXRD analysis were performed on its polycrystalline samples. Based on the re-crystallization experiment, the CABF sample after melting was almost the same as the original one; however, owing to the loss of volatile fluorines, a small amount of Cs₂Al₂(B₃O₆)₂O²⁸ was obtained (Fig. S6a^†). Thermogravimetric (TG) analysis shows that CABF has nearly no weight loss until 900 °C, and the differential scanning calorimetry (DSC) data show two sharp endothermic peaks around 505 and 702 °C on the heating curve (Fig. 2a). In addition, variable temperature PXRD measurements were performed (Fig. S6b^†), demonstrating a possible phase transition of CABF near 500 °C (consistent with the first peak on the DSC curve). Therefore, an excessive amount of fluorides should be used as the flux for crystal growth of CABF. Different from many fluorooxoborates (e.g. CaB₅O₇F₃ and SrB₅O₇F₃ (ref. 29–31)), CABF can survive at relatively high temperature in an open system after replacing BO₃F with AlO₃F units, which is an important advantage for large crystal growth.Open in a separate windowFig. 2(a) Thermal behaviour of CABF. (b) The diffuse reflectance spectrum of CABF. (c) PSHG measurements at 1064 nm. (d) Calculated type I phase-matching condition of CABF. Dashed lines: refractive-indices of fundamental light. Solid lines: refractive-indices of second-harmonic light. The λ_SH of CABF is estimated by satisfying n_Z(ω) = n_X(2ω). The SHG-weighted electron density maps of the occupied (e) and unoccupied (f) states in the VE process.CABF exhibits a wide transparency window from 300 to 1100 nm, with a short UV cut-off edge below 190 nm (Fig. 2b). Even at 190 nm, the reflectance is nearly 40%. DFT calculations of the band gap using the HSE06 hybrid functional with high accuracy result in 7.49 eV (corresponding to 166 nm), further indicating that CABF could be an excellent candidate for optical materials operating in the DUV region. As described earlier,^26,28 the large band gap of CABF could be attributed to the elimination of dangling bonds in [B₃O₆] units and the introduction of fluorine.Powder SHG (PSHG) measurement using the Kurtz–Perry method³² reveals that CABF has SHG intensities of 2.0× KDP (particle size range: 200–250 μm) under 1064 nm fundamental wave laser radiation. It also suggests that CABF is phase matchable as the SHG intensities continue to rise until it attains the maximum (Fig. 2c). Under 532 nm fundamental wave laser radiation, CABF also shows phase matching behavior and has an SHG response about 1/6 times that of BBO in the particle size range of 200–250 μm (Fig. S7^†). The SHG response of CABF is larger than that of KBBF (∼1.2× KDP) and suitable for DUV NLO applications. Applying Kleinmann symmetry in point group mm2, there are three nonzero and independent SHG coefficients, and our calculated SHG coefficients for CABF are d₁₅ = −0.027 pm V⁻¹, d₂₄ = 0.941 pm V⁻¹, and d₃₃ = −0.955 pm V⁻¹, which are consistent with the PSHG results. The SHG ability of CABF is comparable to those of other DUV NLO materials, including CBF (1.9× KDP),²⁶ ABF (3× KDP),²⁵ MB₅O₇F₃ (M = Ca and Sr, 2.3–2.5× KDP),^29–31 (NH₄)₂PO₃F (1.0× KDP),¹² and NaNH₄PO₃F·H₂O (1.1× KDP).¹⁶ Compared with the CBF archetype, the increased SHG response of CABF can be understood by considering the geometry factor of NLO-active [B₃O₆] units. Structurally, CABF is pretty much like CBF; however, the substitution of [BO₃F] by [AlO₃F] results in structural modulation of the [B₃O₆] groups. As shown in Fig. S8,^† the rotation angle (φ) and deviation angle (θ) of the B₃O₆ groups decrease from 33 and 6.8° in CBF, to 25 and 4.4° in CABF, respectively. The smaller φ and θ angles in CABF result in more “coplanar and aligned” [B₃O₆] units, that is, a favorable arrangement for generating the SHG response according to the anionic group approach.⁶The phase-matching ability is a key factor that should never be ignored for any practical DUV NLO application, which relies on both birefringence (Δn) and its dispersion of the crystal.³³ The calculated Δn of CABF is 0.091 at 1064 nm, large enough to satisfy DUV phase-matching (Fig. S10^†). Based on the dispersion of the refractive indices, the shortest SHG phase-matching wavelength (λ_SH) of CABF can reach 182 nm (Fig. 2d), which is comparable to that of KBBF and other recent reported NLO materials (Table S4^†). To the best of our knowledge, the phase-matching wavelength of CABF is the shortest among the SHG-active aluminum borates and aluminum borate fluorides (Table S4^†).To correlate the observed optical properties and electronic structure of CABF, band structure calculations were performed. The partial density of states (PDOS) projected on the constitutional atoms (see Fig. S9^†) demonstrates that B-2p, Al-3p, O-2p, and F-2p states occupy the frontier orbitals (the top of the valence band and the bottom of the conduction band). The high overlap of these orbitals indicates that both the [B₃O₆] and [AlO₃F] groups determine the SHG response and optical birefringence in CABF, while the Cs⁺ cations show a very small contribution. This observation was also supported by the SHG-weighted electron density analysis.³⁴ The directly perceived images (Fig. 2e, f, and S11^†) indicate that the 2p orbitals of O/F atoms in the occupied states and the anti π-orbitals of the [BO₃] groups in the unoccupied states dominate the SHG contribution. In particular, the O or F atoms in [AlO₃F] groups reveal considerable SHG density values, which further suggests that the [AlO₃F] group could be a promising NLO-active unit.