Evidence for Photocatalyst Involvement in Oxidative Additions of Nickel-Catalyzed Carboxylate O-Arylations

Dual photocatalysis and nickel catalysis can effect cross-coupling under mild conditions, but little is known about the in situ kinetics of this class of reactions. We report a comprehensive kinetic examination of a model carboxylate O-arylation, comparing a state-of-the-art homogeneous photocatalyst (Ir(ppy)3) with a competitive heterogeneous photocatalyst (graphitic carbon nitride). Experimental conditions were adjusted such that the nickel catalytic cycle is saturated with excited photocatalyst. This approach was designed to remove the role of the photocatalyst, by which only the intrinsic behaviors of the nickel catalytic cycles are observed. The two reactions did not display identical kinetics. Ir(ppy)3 deactivates the nickel catalytic cycle and creates more dehalogenated side product. Kinetic data for the reaction using Ir(ppy)3 supports a turnover-limiting reductive elimination. Graphitic carbon nitride gave higher selectivity, even at high photocatalyst-to-nickel ratios. The heterogeneous reaction also showed a rate dependence on aryl halide, indicating that oxidative addition plays a role in rate determination. The results argue against the current mechanistic hypothesis, which states that the photocatalyst is only involved to trigger reductive elimination.

Roughly two hours prior to each experiment, the ReactIR console was purged and filled with liquid nitrogen. A background spectrum was recorded shortly before attaching the reaction vessel to the ReactIR probe. To maintain a constant operating temperature during the course of longer experiments, the ReactIR console was replenished with new liquid nitrogen every 12 hours.
The 440 nm lamp was permanently affixed to a metal rod such that the edge of the lamp was 3.5 cm from the middle of the diameter of the ReactIR probe. During our initial studies we found that the greatest source of error derived from inconsistencies in lamp distance and orientation. As such, neither the 440 nm lamp nor ReactIR probe were moved nor disturbed during the months required for data collection.
2.1.2 General experimental procedure S5 A custom-made vial with a sidearm attached (19 x 100 mm, see Figure S1) was equipped with a stir bar and charged with photocatalyst, N-Boc proline, and aryl iodide. Subsequently, DMSO (anhydrous, 3 mL), NiCl2·glyme and dtbbpy from a stock solution, and BIPA were added. Both necks of the vial were sealed with septa and Parafilm. The reaction mixture was sonicated for 5 min followed by stirring for 5-10 min until fine dispersion of the solids was achieved. The flask was then transported to the ReactIR where the larger septum was removed and the vessel immediately attached to the probe.
To ensure an airtight seal, a PTFE adapter was affixed to the probe, to which the vessel was snugly attached. The vessel was continually degassed with Ar for 15 minutes through the sidearm with thin needles. The mixture was stirred for 5 minutes again to re-ensure mixing of the components while data collection started on the ReactIR. After this period the 440 nm lamp was turned on, and this initiation time was marked with the ReactIR proprietary software.

General experimental procedure for delayed injection experiments
A custom-made vial with a sidearm attached (19 x 100 mm, see Figure S1) was equipped with a stir bar and charged with all reaction components except for one. Both necks of the vial were sealed with septa and Parafilm. The reaction mixture was sonicated for 5 min followed by stirring for 5-10 min until fine dispersion of the solids was achieved. The flask was then transported to the ReactIR, where the larger septum was removed and the vessel immediately attached to the probe. To ensure an airtight seal, a PTFE adapter was affixed to the probe, to which the vessel was snugly attached. The vessel was continually degassed with Ar for 15 minutes through the sidearm with thin needles. The mixture was stirred for 5 minutes again to re-ensure mixing of the components while data collection started on the ReactIR. After this period the 440 nm lamp was turned on to 100% power, and this initiation time was marked with the ReactIR proprietary software. After five minutes, the last component (in a DMSO solution) was injected.

Considerations after experiments
To aid in separation of peaks, a negative second derivative function was applied to the raw ReactIR absorbance data. After subtraction of reference spectra, the product peak arrives at ~1764 cm -1 while disappearance of the starting material can be observed at a peak around ~761 cm -1 . Raw data from iCiR was ported to Excel (Microsoft) for processing. Final data were then plotted in Excel (Microsoft) or OriginPro 2015 (OriginLab). S6 Initial concentrations of all starting materials were determined from reaction stoichiometry. Final concentration of the product was determined from 1 H-NMR analysis. This method was validated as described in Section 3.  3. On the validity and reproducibility of data 3.1 Ensuring validity of data Scheme S1. Experiment to independently validate the method.
An experiment following the general procedure outlined (see 2.1.2) above was conducted, according to the stoichiometry in Scheme 1. Notably, added to the normal reaction mixture was 100 mM 1,3,5trimethoxybenzene, the internal standard used for all ex situ NMRs.
Upon initiation of light and periodically thereafter, small aliquots were withdrawn. The timepoints of these aliquots were noted, and each aliquot was analyzed with 1 H-NMR. S9 Previously we have demonstrated that final NMR yield of these reactions is accurately reflected by isolated yield. 2 As such, NMR can be considered a reliable benchmark for comparison of our in situ method.
ReactIR yield was calculated from raw absorbance data that was normalized and scaled, tethered to the final NMR yield. The overlay below between two completely independent methods indicates that ReactIR is a competent measure of reaction progress.

Ensuring reproducibility of data
Frequently experiments were repeated to ensure that data was reproducible. Below are shown one example of such a data set.
In general, for data sets that were to be compared with one another, the same stock solution of NiCl2•glyme and dtbbpy was used. As stated above, the lamp was secured and unmoved during the course of all experiments for this work. Those two factors were paramount for reproducible data collection.
Two of these NMRs are shown below.   Figure S4. Effect of lamp power on reaction speed. In order to best saturate the nickel catalyst with excited photocatalytic species, the highest lamp setting was chosen for "photon-unlimited" experiments in this study.
50% power was used for the "photon-limited" studies due to slightly advantageous yield.  Table S3. Tabulated yields and side products from screening lamp power. S13

Cursory examination of induction period
A custom-made vial with a sidearm attached (19 x 100 mm, see Figure S1) was equipped with a stir bar and charged with all reaction components except for one. Both necks of the vial were sealed with septa and Parafilm. The reaction mixture was sonicated for 5 min followed by stirring for 5-10 min until fine dispersion of the solid photocatalyst was achieved. The flask was then transported to the ReactIR, where the larger septum was removed and the vessel immediately attached to the probe. To ensure an airtight seal, a PTFE adapter was affixed to the probe, to which the vessel was snugly attached. The vessel was continually degassed with Ar for 15 minutes through the sidearm with thin needles. The mixture was stirred for 5 minutes again to re-ensure mixing of the components while data collection started on the ReactIR. After this period the 440 nm lamp was turned on to 100% power, and this initiation time was marked with the ReactIR proprietary software. After five minutes,

Order of catalyst, photon-limited
Following the delayed injection procedure (see section 2.1.3 -Ni•L delayed) described above, experiments were conducted varying only the concentration of nickel and ligand.

Entry
• 3.33 mg/mL g-CN • 6.66 mg/mL g-CN Figure S7. In the photon-limited regime, doubling the amount of photocatalyst has no effect on reaction rate. Delayed injection of NiCl2•glyme and dtbbpy solution was used to assist potential VTNA manipulations, which were not used in this instance. S16

Towards a photon-unlimited regime
Following the general procedure described above (see section 2.1.2), experiments were conducted varying only the concentration of nickel and ligand. Initial rates were determined with a Savitsky-Golay filter at the t = 15 min data point.

Catalyst order
Following the general procedure described above (see section 2.1.2), experiments were conducted varying only the concentration of nickel and ligand.
[3] (Plot shown in Figure 2B.)  Table S16. Tabulated yields and side products from photon-unlimited experiments to determine rate dependence on heterogeneous photocatalyst loading.

Hammett plot
Following the general procedure described above (see section 2.1.2), experiments were conducted varying only the aryl iodide. Initial rates were determined with a Savitsky-Golay filter after the end of the induction period, which varied among halides in this series.
As in situ IR is an integral measurement, 3 obtaining rate data inherently produces lots of noise. As such, finding one 'initial rate' data point for extremely slow reactions is not possible with any reasonable amount of accuracy. For reactions that proceeded at a negligible pace, the initial rate was simply approximated as zero. For example, the most processed or smoothed (Savitsky-Golay filter, 19pt) rate data for the p-CH3 substituent is shown below, from which any chosen value would have little significance.
(Data shown graphically in Figure 4C.)  (Results of the reaction are presented in tabular form in Table 1 and graphical form in Figure 3A.) When normalized [1] vs normalized [t] is plotted, the curvature change between the experiments becomes clearer: Figure S17. Flatter curvature of the lower photocatalytic loading is indicative of lower-order overall kinetics and a higher likelihood that rate is limited by photon-related processes. (mM) Figure S18. VTNA indicates that catalyst is first-order in the photon-limited regime of the homogeneous reaction. [3] (mM)  amIr29  100  0  86  5  5  96  amIr28  50  1  41  3  3  48   Table S19. Tabulated yields and side products from homogeneous photon-limited same excess experiment.   0  87  5  3  95  amIr17  0.10  0  92  4  3  99  amIr16  0.20  0  94  3  4  101   Table S18. Tabulated yields and side products from photon-unlimited experiments assessing catalyst order.

Different excess experiments
Following the general procedure described above (see section 2.1.2), experiments were conducted varying only the concentrations of aryl iodide 1 and carboxylic acid 2.  amIr31  100  100  1  88  4  7  100  amIr33  100  50  22  50  6  19  97  amIr41  50  150  0  47  3  1  51  amIr42  100  200  0  93  5  3  101  amIr43  150  100  26  100  6  16  148   Table S20. Tabulated yields and side products from homogeneous photon-limited different excess experiments.   The purpose of this section was to corroborate the data from Section 7 through finding the linearabsorption (photon-unlimited) regime through the same procedure that was used in the heterogeneous case. These data sets are the same as Section 7 but with a lower [PC] and [Ni]. Agreement between Sections 7 and 8 helps reinforce that the regime is photon-unlimited. The reaction times are longer in this section, but otherwise the trends are the same as Section 7. This remains as a note for researchers that the photon-unlimited regime can be accessed through either procedure.
8.1 Toward a photon-unlimited regime: Initial rate studies, 0.5 mM Ir(ppy)3 Following the general procedure described above (see section 2.1.2), experiments were conducted varying only the concentration of nickel and ligand. Initial rates were determined from product formation with a Savitsky-Golay filter at the t = 10 min data point. Figure S27. Initial rate studies performed to determine region in which nickel catalyst is not limited by transfer from excited photocatalytic species.