Exo-cage catalysis and initiation derived from photo-activating host–guest encapsulation

Coordination cage catalysis has commonly relied on the endogenous binding of substrates, exploiting the cavity microenvironment and spatial constraints to engender increased reactivity or interesting selectivity. Nonetheless, there are issues with this approach, such as the frequent occurrence of product inhibition or the limited applicability to a wide range of substrates and reactions. Here we describe a strategy in which the cage acts as an exogenous catalyst, wherein reactants, intermediates and products remain unbound throughout the course of the catalytic cycle. Instead, the cage is used to alter the properties of a cofactor guest, which then transfers reactivity to the bulk-phase. We have exemplified this approach using photocatalysis, showing that a photoactivated host–guest complex can mediate [4 + 2] cycloadditions and the aza-Henry reaction. Detailed in situ photolysis experiments show that the cage can both act as a photo-initiator and as an on-cycle catalyst where the quantum yield is less than unity.


Materials and Methods
Unless stated otherwise, all reagents and solvents were purchased from Alfa Aesar, VWR, Fluorochem or Sigma Aldrich and used without further purification.Where the use of anhydrous solvent is stated, drying was carried out using a solvent purification system manufactured by Glass Contour.Column chromatography was carried out using Geduran Si60 (40-63 µm) as the stationary phase and TLC was performed on precoated Kieselgel 60 plates (0.20 mm thick, 60F254.Merck, Germany) and observed under UV light at 254 nm or 365 nm.All reactions were carried out under air, unless stated otherwise.
Abbreviations used in the Supporting Information include:  (J) are reported in hertz (Hz).Standard abbreviations indicating multiplicity were used as follows: m = multiplet, q = quartet, t = triplet, d = doublet, s = singlet, bs = broad singlet.All analysis was performed with MestReNova, Version 12.0.3.All assignments were confirmed using a combination of COSY, NOESY, HMBC and HSQC NMR.

RT
All UV/Vis spectroscopy was carried out on a Shimadzu UV-1900 Spectrophotometer running UV Probe, Version 2.70 (Shimadzu).All data was analyzed and plotted using Origin 2015 software.All measurements were made at room temperature (16-21 °C) in CH2Cl2 using a fused silica cuvette with a 10 mm path length, unless stated otherwise.

Ex situ reactions
Aside from Section 5, all ex situ irradiated reactions were carried out under either blue (460 nm) or green (530 nm) LED irradiation.The photoreactor (Figure S1a-c) is composed of a 24 V direct current (DC) power supply and a right angle through hole DC power socket, with external cooling provided through the use of a DC axial fan, purchased from RS Components.For the LED lights, the photoreactor is equipped with either LEDST60BL (blue, 460 nm) or LEDST60GN (green, 530 nm) LED strips purchased from Expert Electrical Supplies Ltd. and connected through a plug-in power supply for LED strips.For aerated experiments, a Penn-Plax AirTech 2K0 pump was used to bubble compressed air through the reaction a rate of 0.3 m/s.
For Section 5, the ex situ irradiated reactions were carried out under irradiation with a 12 W white LED provided by a TACKLIFE Cree 12 W LED torch (Figure S1d).

In situ reactions
The LED-NMR apparatus was constructed in-house and based on the design developed by Gshwind and co-workers. 1 A fibre-coupled LED (λ = 455 nm) powered by a BLS-SA04-US LED driver was purchased from Mightex Systems.This is connected to an Opt 20L TTL trigger box from Hi-Tech Scientific, fitted with a BNC splitter to allow simultaneous communication with the NMR console and the LED driver.The operation of an external trigger switch by the experimenter thus causes the spectrometer to begin recording measurements simultaneously to the reaction beginning to be illuminated.The fibre-optic cable is a FP1500URT purchased from Thor Labs.In order to prepare the fibre for our desired purpose the various layers of cladding (external rubber, followed by Teflon fibres, then a loose plastic sheath, then several layers of adhered polymer) were carefully stripped away from one end with a scalpel to the appropriate length and the final layer of optical coating was removed by dissolution in acetone.The end of the cable was very gently roughened with sandpaper to give uniform illumination in all directions, placed within a quartz coaxial insert and permanently attached using epoxy resin.All experiments are conducted within amberised standard 5 mm NMR tubes containing the fibreoptic cable held within the quartz insert to give a total reaction volume of 280 µL.Amberised NMR tubes and quartz coaxial inserts were both purchased from Norrell Scientific.
All LED-NMR experiments were performed on a Bruker Avance III 400 MHz spectrometer equipped with a Prodigy cryoprobe.The reactions were performed in CD2Cl2 and monitored by 1 H NMR spectroscopy, using a series of single scan spectra (zg30, acquisition time = 2 s) with a set delay between them.In order to ensure accurate measurement of the very early periods of the reaction, the pulse program was set up such that the first spectrum was acquired simultaneously to the irradiation of the sample beginning.To ensure that sufficient time had passed to allow the sample to fully relax between NMR excitations the delay between individual single-scan spectra was never shorter than 5 s.

Control Reactions
To ensure all components were essential for reactivity, a variety of different control reactions were carried out.These followed the above procedure and included the following changes:  Omitting light (amberised NMR tube). Omitting Q1.  Omitting C1.Where species were omitted, stock solution volumes were replaced with CD2Cl2.

Product Identification
The product of each NMR scale reaction was identified (and thereby yields calculated) by comparing the spectra to previously reported 1 H NMR spectroscopic data. 6

530 nm irradiation
The reaction was irradiated with 530 nm light for 5 min, followed by 10 min (15 min total), followed by 45 min (1 h total).The reaction was monitored by recording 1 H NMR spectra before the addition of C1 and after each irradiation session.The reaction was irradiated with 460 nm light for 5 s, 30 s or 1 min.The reaction was monitored by recording 1 H NMR spectra before the addition of C1 and after irradiation.

Lower Catalyst Loadings
The reactions with lower catalyst loadings (0.75, 0.5 and 0.25 mol%) were prepared as above, reducing the volumes of Q1 and C1 used as necessary, replacing the reduced volume of C1 stock with CD2Cl2 to ensure constant overall volume.
For each NMR reaction, the solution (280 µL) was added to an amberised NMR tube, into which the coaxial insert containing the fibre-optic cable was inserted and the tube was sealed.This was lowered into the spectrometer probe and the sample was locked, tuned/matched and shimmed.An initial reference spectrum was taken prior to illumination of the sample.
Initial experimentation using 5 mol% Q1⸦C1 loading and the maximum possible LED intensity ('100% power') revealed that the reaction was too fast to conveniently monitor, and so it was slowed down by dropping the catalyst loading to 1 mol% and standard LED intensity to 5% of the maximum.Between each reaction the quartz/fibre optic insert was thoroughly cleaned by rinsing with acetone and wiping with a paper towel.

Orders in light intensity and catalyst loading
Light intensity studies were conducted by varying the LED intensity and catalyst loading studies were conducted by varying the catalyst concentration of the solution.Initial rates were taken by approximating the first 40 s (5 data points for catalyst loading studies) or first 30 s (4 data points for light intensity studies) of the reactions as a straight line. 3.6.

On/off studies
Using the programming functions of the LED driver, the reaction (with 1 mol% catalyst) was illuminated at 455 nm and 5% intensity for 60 s followed by 300 s of darkness.This process was repeated twice, and the reaction was then illuminated for a further 120 s prior to the light being turned off, although the reaction was monitored for a further 1000 s (as shown in Figure S15).

Measurement of the reaction kinetics at standard catalyst loadings and light intensity
Following initial measurements that showed that the standard reaction was over within at least 10 s under LED-NMR conditions at maximum light intensity, an interleaving strategy was used to measure the form of the reaction kinetics under these conditions.A large batch of identical reaction solution at 5 mol% catalyst loading was prepared, as discussed above.A series of single-scan NMR measurements were then made, using different delays (0-5 s) between the initiation of illumination and measurement of the first data point in each fresh 280 µL sample.Each of these kinetic traces was then plotted on the same axis to give a full kinetic picture.Two reactions were prepared identically, with the exception that each contained 1 mol% of C1 synthesised in two distinct batches.They were both run under the same conditions with in situ monitoring and the resulting overlaid kinetic profiles can be seen below: 3.6.6Comparison to existing photocatalyst Ru(bpz)3(BArF)2 A solution consisting of 1 mol% (0.1 mM) Ru(bpz)3(BArF)2 was also prepared from stock solutions following a similar general procedure to that discussed previously, and the kinetics of this reaction under irradiation at 5% light intensity was measured using two distinct batches of Ru(bpz)3(BArF)2 under otherwise identical conditions.3.6.7 Addition of NaBArF to the reactions A 280 µL aliquot was taken from a fully-assembled reaction solution.To the remainder of the solution was added sufficient NaBArF to saturate the solution.A 280 µL aliquot of this solution was placed in an amberised NMR tube and the reaction was monitored by LED-NMR.Comparison between the NaBArFsaturated solution and the solution without additional NaBArF was conducted using both 1 mol% Q1⸦C1 and 1 mol% Ru(bpz)3BArF2.A reaction solution was prepared containing 1 mol% Q1.This solution was saturated with NaBArF and the resulting reactivity was monitored by LED-NMR illumination at 455 nm and 5% LED intensity.This was compared to an identical solution with no addition of NaBArF, where no reactivity was observed.Using 5 mol% of NaBArF with no photocatalyst or Q1 yielded no reactivity.Saturating a solution of only 1 and 2 in CD2Cl2 with NaBArF yielded trace amounts of product after 15 minutes of irradiation at 455 nm (see Figure S23).8.1 8.0 7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 7.1 7.0 6.9 6.8 6.7 6.6 6.5 6.4 6.3 6.2 6.1 6.0 5.9 5.8 S24 3.6.9 1 H NMR spectra of in situ irradiated [4+2] cycloadditions All of the below manipulations were conducted in a dark room under red illumination, and the experiment was designed following literature precedent. 7 Sulfuric acid (14 mL, 17.8 M) was diluted up to 500 mL with H2O to form a 0.5 M aqueous solution of sulfuric acid.Sulfuric acid (10 mL, 0.5 M) was diluted up to 100 mL with H2O to form a 0.05 M aqueous solution of sulfuric acid.Potassium ferrioxalate trihydrate (1.85 g, 3.76 mmol) was dissolved in sulfuric acid (25 mL, 0.05 M).Sodium acetate (22.6 g, 277 mmol, 49 eq.) was dissolved in sulfuric acid (100 mL, 0.5M).1,10-phenanthroline (1.02 g, 5.67 mmol. 1 eq.) was then added.
An aliquot of the potassium ferrioxalate solution (0.28 mL) was added to an amberised NMR tube, into which was then placed the LED-NMR insert.The solution was then illuminated for the required time at 455 nm, at the 5% setting on the LED driver.10 µL of this solution was then added to an aliquot of the sodium acetate:phenanthroline solution (5 mL) containing a magnetic stirrer bar.The mixture was left to stir for one hour in the dark.A UV-Vis spectrum was then taken of this solution and the absorbance at 510 nm was used to calculate the amount of Fe 2+ in the illuminated solution, as shown below: From the Beer-Lambert law:  By least squares fitting the gradient of the graph is calculated to be 1.6 ± 0.02 × 10 −9 mol s −1 of Fe 2+ generated.
= path length (measured to be 0.44 mm for our LED-NMR apparatus) From the literature, logε of the K3[Fe(C2O4)3] at 450 nm is approximately 1.3. 9Extrapolation of the graph to 455 nm puts this at approximately 1.2 and we take this as a reasonable approximate of the true value.Stock solutions were prepared using spectrophotometry grade CH2Cl2 and standard volumetric glassware.Solutions of the host-guest complex Q1⸦C1 were prepared in a 1:1.2 ratio to minimise the concentration of unbound quinone.Stock solutions were diluted and transferred to quartz cuvettes with 1 mm pathlength for spectral acquisition.
UV spectra were recorded on Ocean-Optics Flame CCD Spectrometer connected via optical fibre to a DH2000-BALUV lamp.The samples were housed in a thermostatted (20.0°C ± 0.1°C) cuvette holder.Acquisition was controlled by the Kinetic Studio software package (version 5).Data were processed and plotted using Excel (replotted using Origin).

Calculating how much product is produced by the active catalyst
For Q1⸦C1: A line was fitted to the graph of Q1⸦C1 loading against initial rate.Each data point represents the average of two different runs.By least squares fitting the gradient was determined to be 1.14 ± 0.03 mM s −1 per mM catalyst.Thus, within this concentration range, for each unit of Q1⸦C1 complex present in solution, we generate 1.14 units of product s −1 .
Taking a reaction with 0.1 mM catalyst loading as a case study: there is 280 µL of a solution containing 0.1 mM Q1⸦C1, i.e. 2.8 × 10 −8 mol of catalyst.Thus, from the above ratio, we produce 3.19 ± 0.09 × 10 −8 mol product s −1 .
For Ru(bpz)3(BArF)2: A line was fitted to the graph of Ru(bpz)3(BArF)2 loading against initial rate.Each data point represents the average of two different runs.By least squares fitting the gradient was determined to be 1.79 ± 0.09 mM s −1 per mM catalyst.Thus, within this concentration range, for each unit of Ru(bpz)3(BArF)2 complex present in solution, we generate 1.79 units of product s −1 .

Calculation of catalyst quantum yield
For Q1⸦C1: Catalyst concentration / mM y = 1.794x+0.0028R 2 = 0.993 For Ru(bpz)3(BArF)2: ϕ = (product) (photons) = 5.0 × 10 −8   −1 6.2 × 10 −10   −1 = 80 ± 10 For the Ru(bpz)3(BArF)2 complex, this value is approximately double that reported by Cismesia and  Yoon. 10 This discrepancy presumably arises from the differences between the two methods used to estimate quantum yield.In this work we estimate it based on an initial rates relationship; in contrast, Cismesia and Yoon use the total yield after a certain amount of time.These two different methods of calculating quantum yield will inevitably result in somewhat different numberscalculations based on end points will typically provide smaller values as addition of some amount of photons to a reaction that is near or at completion will result in much less (or no) product formation than addition of that amount of photons near the beginning of a reaction.In contrast, calculations based on initial rates trend towards giving a higher value as they use a regime where the maximum proportion of absorbed photons are productively utilised to generate product.The latter is representative of the optimal radical chain length; however, it does not take into account progressive inhibition of the reaction, catalyst decomposition and other more complex mechanistic phenomena.
The reaction was screened with the following solvents:

Additive Experiment
The above procedure for CD2Cl2 was modified with the addition of NBu4•PF6 (0.06 mg, 1.5 × 10 −4 mmol, 0.025 eq.) prior to collecting the initial 1 H NMR spectrum (i.e.prior to the addition of C1).
In the absence of light, C1 (7.2 mg, 1.5 × 10 −3 mmol, 0.05 eq.) was added and the microwave vial was sealed.An aliquot (400 µL) was immediately taken and added to an NMR tube containing CD2Cl2 (200 µL), which was kept in the dark to avoid photoreaction.The reaction was irradiated with 460 nm light for 2 min under a constant stream of compressed air.Another aliquot (400 µL) was taken and added to an NMR tube containing CD2Cl2 (200 µL).Product yields were calculated by comparing integrals of the starting material and product to that of the internal standard.

Product Identification
The product of each NMR scale reaction was identified (and thereby yields calculated) by comparing the spectra to previously reported 1 H NMR spectroscopic data.Stock solutions in CD2Cl2 were prepared for: 2-fluoromesitylene (internal standard, 100 mM), 1,3cyclohexadiene (80 mM), C1 (5 mM) and Ru(bpz)3(BArF)2 (5 mM).A stock solution in CH2Cl2 was prepared for Q1 (0.5 mM) and the appropriate amount was measured into a 1 mL volumetric flask.The CH2Cl2 was then evaporated and the appropriate volumes of all stock solutions were measured into the flask, with the remaining volume then being made up to 1 mL.The solution was thoroughly homogenised before a 280 µL aliquot was taken for each in situ monitoring experiment.

Variation of Q1⸦C1 concentration
Five reactions were prepared with varying concentrations of Q1⸦C1 (0.25-0.063mM).The effect of varying Q1⸦C1 loading on rate of consumption of cyclohexadiene (Figure S37) and formation of endodimer (Figure S38).Each co-plotted separately for ease of visualisation.

Variation of light intensity
Two identical solution aliquots were irradiated at 455 nm, at 5% and 100% of the maximum LED intensity.It was observed that raising the light intensity raised the reaction rate, but did not impact on the overall kinetic profile.4.4.7 Calculation of reaction quantum yield, using Q1⸦C1 as the catalyst The same procedure as in Section 3.7 was followed, using the previously measured light intensity and Q1⸦C1 extinction coefficient.It was previously shown that 0.1 mM catalyst will absorb 1.3 × 10 −9 mol photons per second.A line was fitted to the graph of Q1⸦C1 loading against initial rate of product formation.By least squares fitting the gradient was determined to be 0.022 ± 0.004 mM s −1 per mM catalyst.
Thus, within this concentration range, for each unit of Q1⸦C1 complex present in solution, we generate 0.022 units of product s −1 .
Taking a reaction with 0.1 mM catalyst loading as a case study: there is 280 µL of a solution containing 0.1 mM Q1⸦C1, i.e. 2.8 × 10 −8 mol of catalyst.Thus, from the above ratio, we produce 6 ± 1 × 10 −10 mol product s −1 .ϕ = (product) (photons) = 6 × 10 −10   −1 1.3 × 10 −9   −1 = 0.5 ± 0.2 Individual analysis of each reaction using the amount of product formed and photons absorbed after a certain time gave values in line with this result.For each reaction, this was calculated for ten individual data points after the reaction had plateaued.
Averaging this value across four different reactions catalysed by Q1⸦C1 yielded a value of 8 ± 1.

Figure S1 :
Figure S1: (a) Schematic representation of the photoreactor shown in (b).(c) Picture displaying the photoreactor in use with a LEDST60GN (green, 530 nm) LED strip.(d) Experimental set up for the Cage-Quinone Catalyzed Aza-Henry Reaction (Section 5) using a 12 W LED torch.

Figure S8 :
Figure S8: 1 H NMR spectrum (500 MHz, CDCl3, 300 K) for the cycloadduct isolated from the reaction of trans-anethole and isoprene in the presence of C1 and Q1 and 460 nm light.

Figure S13 :
Figure S13: (a) Concentration vs time plots of consumption of trans-anethole (circles) and production of cycloadduct (triangles) using 5 mol% Q1⸦C1 in CD2Cl2 using different LED intensities: (i) 10% LED intensity (purple circles and triangles); (ii) 5% LED intensity (blue circles and triangles); (iii) 1% LED intensity (yellow circles and triangles).(b) Linear fittings of initial points in graph (a) to yield initial rates (mM s −1 ).(c) Initial rates obtained from the linear fittings of (b).(d) Initial rate of product formation vs LED intensity plot shows a linear relationship.

Figure
Figure S14: 1 H NMR spectra (400 MHz, CD2Cl2, 300 K) for the reaction of trans-anethole (10 mM) and isoprene (400 mM) in the presence of C1 (0.1 mM) and Q1 (0.1 mM) irradiated in situ with 455 nm light (5% power).Periods of irradiation indicated by the blue bar, periods of darkness indicated by the navy bar.trans-Anethole is highlighted in red, product is highlighted in blue, overlapping trans-anethole and cycloadduct peaks highlighted in purple.δ H / ppm

Figure S16 :
Figure S16: Concentration vs time plot of consumption of trans-anethole (red symbols) and production of cycloadduct (blue symbols) using 5 mol% Q1⸦C1 in CD2Cl2 irradiated with 455 nm light (100%).Different data sets are shown using different symbols.

Figure S18 :
Figure S18: Overlaid kinetic plots of two identical reactions each using 1 mol% of a different batch of C1 of the conversion of trans-anethole (circles) to product (triangles).

Figure S19 :
Figure S19:Overlaid kinetic plots of two identical reactions each using 1 mol% of a different batch of Ru(bpz)3(BArF)2 of the conversion of trans-anethole (circles) to product (triangles).

Figure S20 :
Figure S20:Overlaid kinetic plots of two otherwise-identical reactions each using 1 mol% Q1⸦C1, where one solution is saturated with NaBArF, of the conversion of trans-anethole (circles) to product (triangles).

Figure S22 :
Figure S22: Overlaid kinetic plots of two otherwise-identical reactions each using 5 mol% Q1 and no additional photocatalyst, where one solution is saturated with NaBArF, of the conversion of trans-anethole (circles) to product (triangles).

Figure S23 :
Figure S23: Partial 1 H NMR spectra (400 MHz, CD2Cl2, 300 K) for the reaction of trans-anethole (10 mM) and isoprene (400 mM) in the presence of NaBArF (saturated solution) irradiated with 455 nm light (15 min), demonstrating formation of trace product only.A selected signal of trans-anethole is highlighted in red and one of product is highlighted in blue.

Figure S25 :
Figure S25: A plot of Fe 2+ generated as a function of LED irradiation time within the NMR tube.

Figure S27 :Figure S28 :
Figure S27: Experimental (points) and simulated (filled line) absorbance at 455 nm of Q1⸦C1 as a function of concentration.The molar absorptivity (inset) is calculated via linear regression, according to the Beer-Lambert Law.

Figure S29 :
Figure S29: Experimental (points) and simulated (filled line) absorbance at 455 nm of Ru(bpz)3(BArF)2 as a function of concentration.The molar absorptivity (inset) is calculated via linear regression, according to the Beer-Lambert Law.

Figure S30 :
Figure S30: Initial rates of product formation against Q1⸦C1 concentration.

Figure S38 :
Figure S38: Kinetic traces of formation of endo-dimer at varying loadings of Q1⸦C1.Irradiation is at 5% LED power at 455 nm.

Figure S44 :
Figure S44: Initial rates of product formation against Q1⸦C1 concentration.

Figure S49 :
Figure S49: UV-vis spectra (CH2Cl2, 298 K) of (i) black line = Q1 (0.5 mM); (ii) purple line = Q1 (0.5 mM) in sat.NaBArF.From this spectrum, the extinction coefficient of Q1 in a saturated CH2Cl2 solution of NaBArF was estimated at 874 M −1 cm −1 Unless stated otherwise, all 1 H, 13 C and 19 F NMR spectra were recorded on either a 500 MHz Bruker AV III equipped with a DCH cryoprobe, a 500 MHz Bruker AV IIIHD equipped with a Prodigy cryoprobe or a 400 MHz Bruker AV III equipped with BBFO+ probe at a constant temperature of 300 K.Chemical shifts are reported in parts per million.Coupling constants Addition of NaBArF to the reactionsReaction solutions were prepared as per the general procedure discussed in Section 4.4.1.In the instances where NaBArF was added, the fully-assembled solution was saturated with NaBArF prior to irradiation.Overlaid kinetic profiles of irradiation of a solution with 5 mol% Q1 as the photocatalyst.The NMR signal corresponding to cyclohexadiene is represented as circles, the NMR signal corresponding to endo-dimer as triangles.No NMR signal corresponding to the exo-dimer is observed.One solution is additionally saturated with NaBArF.