Nickel-mediated N–N bond formation and N2O liberation via nitrogen oxyanion reduction

The syntheses of (DIM)Ni(NO3)2 and (DIM)Ni(NO2)2, where DIM is a 1,4-diazadiene bidentate donor, are reported to enable testing of bis boryl reduced N-heterocycles for their ability to carry out stepwise deoxygenation of coordinated nitrate and nitrite, forming O(Bpin)2. Single deoxygenation of (DIM)Ni(NO2)2 yields the tetrahedral complex (DIM)Ni(NO)(ONO), with a linear nitrosyl and κ1-ONO. Further deoxygenation of (DIM)Ni(NO)(ONO) results in the formation of dimeric [(DIM)Ni(NO)]2, where the dimer is linked through a Ni–Ni bond. The lost reduced nitrogen byproduct is shown to be N2O, indicating N–N bond formation in the course of the reaction. Isotopic labelling studies establish that the N–N bond of N2O is formed in a bimetallic Ni2 intermediate and that the two nitrogen atoms of (DIM)Ni(NO)(ONO) become symmetry equivalent prior to N–N bond formation. The [(DIM)Ni(NO)]2 dimer is susceptible to oxidation by AgX (X = NO3−, NO2−, and OTf−) as well as nitric oxide, the latter of which undergoes nitric oxide disproportionation to yield N2O and (DIM)Ni(NO)(ONO). We show that the first step in the deoxygenation of (DIM)Ni(NO)(ONO) to liberate N2O is outer sphere electron transfer, providing insight into the organic reductants employed for deoxygenation. Lastly, we show that at elevated temperatures, deoxygenation is accompanied by loss of DIM to form either pyrazine or bipyridine bridged polymers, with retention of a BpinO− bridging ligand.

General. All manipulations were carried out under an atmosphere of ultra-high purity nitrogen using standard Schlenk techniques or in a glovebox under N 2 . Solvents were purchased from commercial sources, purified using Innovative Technology SPS-400 PureSolv solvent system or by distilling from conventional drying agents and degassed by the freeze-pump-thaw method thrice prior to use. Glassware was oven-dried at 150 °C overnight and flame dried prior to use. Synthesis of (DIM)NiBr 2 .

N N Ni
Br Br DIM (0.705 g, 2.22 mmol) was dissolved in DCM (5 mL) and added to a slurry of (DME)NiBr 2 (DME = dimethoxyethane, 0.684 g, 2.22 mmol) in 5 mL of DCM. This caused a color change to deep orange in 20 minutes. The reaction was stirred for 12 hours with no further color changes. All volatiles were removed under reduced pressure, giving rise to a deep orange solid, which was washed with 3 x 5mL diethyl ether. A crude orange powder was isolated in 82% yield.

Synthesis of (DIM)Ni(NO
In a 20 mL scintillation vial, (DIM)NiBr 2 (0.0867 g, 0.161 mmol) was dissolved in 5 mL of acetonitrile. To this solution, silver nitrate (0.0574 g, 0.338 mmol, dissolved in 2 mL acetonitrile) was added dropwise, causing an immediate color change from orange to yellow-brown with immense precipitation of silver bromide. The reaction was stirred for 30 minutes with no additional color changes, upon which AgBr was filtered off giving rise to a homogenous yellow-brown solution. All volatiles were removed to give a tan powder, which was then redissolved in THF and filtered through a plug of celite. Removal of volatile material under reduced pressure gave rise to a pale yellow powder, isolated in 76% yield. Crystals suitable for single crystal X-ray diffraction were grown by diffusion of pentane vapors into a concentrated DCM solution. In a 20 mL scintillation vial, (DIM)NiBr 2 (0.0431 g, 0.080 mmol) was dissolved in 5 mL of acetonitrile. To this solution, silver nitrite (0.0258 g, 0.168 mmol, dissolved in 2 mL acetonitrile) was added dropwise, causing a slight color change to dark orange and immediate precipitation of byproduct silver bromide. The reaction was stirred for 30 minutes with no additional color changes, upon which AgBr was filtered off giving rise to a homogenous yellow-brown solution. All volatiles were removed to give a yellow powder, which was then redissolved in THF and filtered through a plug of celite. Removal of volatile material under reduced pressure gave rise to a dark yellow powder, isolated in 84% yield. Crystals suitable for single crystal X-ray diffraction were grown by diffusion of pentane vapors into a concentrated DCM solution.
Synthesis of (DIM)Ni( 15 NO 2 ) 2 . In a 20 mL scintillation vial, (DIM)NiBr 2 was dissolved in 5 mL of dichloromethane. To this stirring solution was added Na 15 NO 2 as a slurry in dichloromethane, as well as a catalytic amount of 18-crown-6 to help solubilize the sodium nitrite. After stirring for 48 hours, the resultant brown solution was filtered through a celite plug to remove the NaBr formed during the course of the metathesis. The filtrate was dried in vacuo to afford a yellow/brown solid, which was washed with pentane (2 x 3 mL) and ether (2 x 3 mL). The washed product was spectroscopically pure by 1 H NMR. To a J-Young tube containing (DIM)Ni(NO 2 ) 2 (0.032 g, 0.068 mmol) in d 8 -THF was added (Bpin) 2 Pz (0.023 g, 0.068 mmol) also dissolved in d 8 -THF. After heating for 6 hours at 45 °C, all starting material was consumed and one, new diamagnetic product was formed. The reaction was accompanied by a slight color change from brown to brown-green. All volatiles were removed in vacuo, and the green solid was rinsed with pentane (3 x 5 mL) to afford a green powder in 80% yield. Crystals suitable for single X-ray diffraction were grown by diffusion of pentane vapors into a concentrated THF solution. To a J-Young tube containing (DIM)Ni(NO)(NO 2 ) (0.027 g, 0.06 mmol) in d 8 -THF was added (Bpin) 2 Pz (0.039 g, 0.12 mmol) also dissolved in d 8 -THF. After spinning for 12 hours at room temperature, all starting material was consumed and free pyrazine and boryl ether are observed by 1 H NMR. The product crystallizes while being agitated in an NMR tube. The crystalline product was isolated in 75% yield by pouring out the mother liquor, and rinsing the dark solid with pentane and ether. [(DIM)Ni(NO)] 2 is remarkably insoluble, and crashes out as crystalline product when the reaction is performed in dichloromethane, acetonitrile, and toluene as well. Independent X-ray crystal structures were obtained from a THF reaction and a DCM reaction, of which the DCM reaction structure gives better data and is the one reported in this manuscript.

ON OTf
A slurry of freshly prepared [(DIM)Ni(NO)] 2 in THF was filtered through a celite plug, upon which the [(DIM)Ni(NO)] 2 precipitate stayed on top of the celite pad. THF (2 x 3 mL) and acetonitrile (2 x 3 mL) were pushed through the celite plug to wash any leftover organic byproducts from the synthesis of [(DIM)Ni(NO)] 2 . The THF and acetonitrile washings filtered through the celite plug were colorless. An equimolar solution of AgOTf in 5 mL of acetonitrile was prepared, the AgOTf solution was filtered through [(DIM)NiNO] 2 on the celite plug, and the resulting filtrate was green. The reduced silver was seen on the top of the celite plug, and the filtrate was dried in vacuo. The resultant green solid was rinsed with pentane (2 x 3 mL), and crystallized in 71% yield from vapor diffusion of pentane into a concentrated THF solution containing (DIM)Ni(NO)(OTf). A sample of [(DIM)Ni(NO)] 2 was isolated on top of a celite plug in the same way reported above. An equimolar solution of AgNO 3 in 5 mL of acetonitrile was prepared, the AgNO 3 solution was filtered through the celite plug carrying [(DIM)NiNO] 2 , and the resulting filtrate was light brown. The oxidized silver was seen on the top of the celite plug, and the filtrate was dried in vacuo to afford a green solid. The solid was rinsed with both pentane (2 x 3 mL) and ether (2 x 3 mL), and redissolved in a minimal amount of THF. To this concentrated solution was added pentane to precipitate a fine green powder in 68% yield.  2 was isolated on top of a celite plug as reported above. An equimolar solution of Ag 15 NO 2 in 5 mL of acetonitrile was prepared, the Ag 15 NO 2 solution was filtered through the celite plug, and the resulting filtrate was light brown. The oxidized silver was seen on the top of the celite plug, and the filtrate was dried in vacuo to afford a green solid. The solid was rinsed with both pentane (2 x 3 mL) and ether (2 x 3 mL), and redissolved in a minimal amount of THF. To this concentrated solution was added pentane to precipitate a fine green powder in 84% yield.
The synthesis of the other isotopomer, (DIM)Ni( 15 NO)(ONO), followed the same experimental procedure as the synthesis of (DIM)Ni(NO)(O 15 NO). However, the incorporation of 15 N at the nitrosyl position requires the starting dimer to be isotopically labelled: To the isotopically labelled dimer was added equimolar unlabelled AgNO 2 to afford the corresponding (DIM)Ni( 15 NO)(ONO) compound in 79% yield. To a J-Young tube containing (DIM)Ni(NO 2 ) 2 (0.035 g, 0.074 mmol) in d 8 -THF was added (Bpin) 2 Pz (0.049 g, 0.148 mmol) also dissolved in d 8 -THF. After heating for 3 hours at 80 °C, all starting material was consumed, and the solution changed color from brown to blue. Upon cooling to room temperature, blue crystalline precipitate forms in the J-Young tube. The solution is subsequently filtered in the glovebox through celite, upon which the crystalline precipitate stays at the top of the celite. The product was rinsed with pentane and ether through the celite to ensure removal of all organic byproducts. The crystalline product was washed through the celite using dichloromethane, and all volatiles were removed to afford a blue solid. This blue solid was washed again with pentane (3 x 5 mL), to afford the product as a blue solid isolated in 73% yield. To a J-Young tube containing (DIM)Ni(NO 2 ) 2 (0.035 g, 0.074 mmol) in d 8 -THF was added (Bpin) 2 Bpy (0.069 g, 0.148 mmol) also dissolved in d 8 -THF. After spinning for 5 hours at room temperature, all starting material was consumed, and the solution changed color from brown to blue. After taking the NMR tube off the spinner blue crystalline precipitate forms in the J-Young tube. The solution is subsequently filtered in the glovebox through celite, upon which the crystalline precipitate stays at the top of the celite. The product was rinsed with pentane and ether through the celite to ensure removal of all organic byproducts. The low solubility of the bipyridine based polymer precluded the collection of a 1 H NMR spectrum.

Experimental procedure for detection of nitrous oxide after (OBPin) 2 Ni 2 (NO) 2 (Bpy) and (OBPin) 2 Ni 2 (NO) 2 (Pz) formation.
Both (OBPin) 2 Ni 2 (NO) 2 (Pz) and (OBPin) 2 Ni 2 (NO) 2 (Bpy) were synthesized according to the established experimental procedure above, but the reactions were carried out in a sealed Schlenk flask to allow for accumulation of gaseous byproducts. Once the polymeric compounds had formed, the Schlenk flask was attached to a gas manifold, which also contained an evacuated gas IR cell. The entire gas manifold was evacuated except the headspace of the reaction flask. The manifold was closed from the vacuum, and a condensation arm built on the side of the gas IR cell was placed in liquid nitrogen. The reaction flask was opened to the static vacuum, allowing any gaseous byproducts formed during the reaction to condense into the gas IR cell for subsequent spectroscopy.
Experimental procedure for detection of nitrous oxide using gas IR following deoxygenation of (DIM)Ni(NO)(ONO). The deoxygenation of (DIM)Ni(NO)(ONO) was performed using (Bpin) 2 Pz in a sealed Schlenk flask, according to the experimental procedure above. The solution was stirred overnight at room temperature to ensure complete consumption of starting materials. Once the reaction was complete, the Schlenk flask containing the reaction mixture was attached to a gas manifold, and the transfer of gaseous byproducts into the gas IR cell was performed according to the general procedure outlined above.
Experimental procedure for detection of nitrous oxide using solution IR following deoxygenation of (DIM)Ni(NO)(ONO). To a solution of (DIM)Ni(NO)(ONO) dissolved in THF in a scintillation vial was added (Bpin) 2 (Bpy), also dissolved in THF. This solution was mixed using a pipette 5 times, before being quickly transferred to a solution IR cell. The solution was then periodically monitored by IR, showing N 2 O formation at the earliest scan, with little change in the intensity of the N 2 O bands during the course of the reaction due to saturation of the THF solution.

Experimental procedure for oxidation of (DIM)Ni(NO)(X) (X = NO 2ˉ or NO 3ˉ) and detection of liberated nitric oxide.
To an acetonitrile solution of either (DIM)Ni(NO)(ONO) or (DIM)Ni(NO)(ONO 2 ) was added AgNO 2 or AgNO 3 , respectively, also in acetonitrile in a 1:1 mole ratio. This solution was placed in a sealed Schlenk flask, and the resultant solution was stirred at room temperature overnight. The nitric oxide can be detected via vacuum transfer of the volatiles to a gas IR cell as outlined by the general procedure above. After detection of nitric oxide, the reaction solutions are dried in vacuo to afford a brown/purple solid. The solids were washed with pentane (2 x 3 mL) and ether (2 x 3 mL) to afford spectroscopically pure (DIM)Ni(NO 2 ) 2 or (DIM)Ni(NO 3 ) 2 as judged by 1 H NMR spectroscopy.

Experimental procedure for nitric oxide disproportionation with a slurry of [(DIM)Ni(NO)] 2 .
A sample of dimeric [(DIM)Ni(NO)] 2 was isolated and suspended as a slurry in THF. This slurry was placed in a Schlenk flask and transferred to the gas manifold. After two freeze-pump-thaw cycles, excess nitric oxide was condensed into the Schlenk flask containing the frozen slurry placed in a liquid nitrogen bath. After the addition of nitric oxide and upon warming to room temperature, the slurry became a homogenous green/brown solution. After stirring for 30 minutes, the volatiles from the reaction flask were vacuum transferred to a gas IR cell according to the general procedure outlined above, showing N 2 O formation as well as excess nitric oxide. The reaction solution was dried in vacuo, and the green solid was washed with pentane (2 x 3 mL) and ether (2 x 3 mL) to afford spectroscopically pure (DIM)Ni(NO)(ONO) as judged by 1 H NMR spectroscopy and the diagnostic ν NO stretching frequency.

Experimental procedure for nitric oxide disproportionation with [(DIM)Ni(NO)] 2 as a solid powder.
A sample of dimeric [(DIM)Ni(NO)] 2 was isolated and mixed with dry KBr. The dimer/KBr solid mixture was transferred to a Schlenk flask and transferred to the gas manifold. After evacuation of the Schlenk flask, two equivalents of nitric oxide were condensed into the Schlenk flask placed in a liquid nitrogen bath. After warming to room temperature and allowing the flask to stay at room temperature for 20 minutes, the Schlenk flask was evacuated, the flask was transferred into the glovebox, and a sample of the dimer/KBr mixture that had been exposed to nitric oxide was used to make a KBr disc for IR spectroscopy. The Schlenk flask was returned to the gas manifold and two more equivalents of nitric oxide were condensed into the Schlenk flask. 30 minutes after the addition of the third and fourth equivalents of nitric oxide, the Schlenk flask was once again evacuated and brought back into the glovebox, upon which an aliquot of the KBr mixture was used for collection of a second IR spectrum. The process of NO addition and assay by solid state infrared spectroscopy was repeated two more times (4 times total to give an 8:1 mole ratio of NO to dimer). No intermediates were seen by solid state IR spectroscopy, and the only diagnostic change observed was the growth of ν NO for (DIM)Ni(NO)(ONO).  : Variable temperature 1 H NMR of (DIM)Ni(NO 2 ) 2 . Upon cooling, the signals for the ortho -CH 3 on the mesityl arm and the meta C-H proton (as judged by relative integrations) both shift downfield; however, the -CH 3 has a larger shift. This leads to accidental degeneracy, best seen at -20 °C. Further cooling alleviates this accidental degeneracy, but also begins to split the ortho -CH 3 and meta C-H signals into two signals each. This can be attributed to decreased rapid rotation of the mesityl groups at lower temperatures, leading to inequivalent ortho -CH 3 and meta C-H groups on each mesityl arm. The lack of rapid rotation does not affect the para -CH 3 groups, or the -CH 3 on the DIM backbone, so these signals do not split. This same pattern is observed for the bis-nitrate complex; however, it is more well resolved in the bis-nitrite.         In spite of being polymeric, Ni 2 (NO) 2 (OBpin) 2 (Pz) dissolves in THF, DCM and MeCN. The 1 H NMR spectrum in CD 2 Cl 2 shows one signal for coordinated pyrazine and one signal for the pinacolate methyls, but the pyrazine signal is unusually broad and suggests some dynamic character. The simplest explanation is that the dissolved species is the repeat unit, Ni 2 (NO) 2 (OBpin) 2 (κ 1 -Pz), and that some dynamic process averages the two inequivalent pyrazine chemical shifts for monodentate pyrazine coordinated to only one end of the Ni 2 (NO) 2 (OBpin) 2 unit, which therefore has one three-coordinate and one four-coordinate nickel. We suggest that either a dissociative or an associate process causes exchange of free and coordinated pyrazine nitrogen, and this process occurs at an intermediate exchange rate under our measurement conditions. Figure S28. 1 Figure S41. Vacuum transfer of the volatiles from the reaction of (DIM)Ni(NO 2 ) 2 with (Bpin) 2 Pz in a 2:3 mole ratio at 80 °C. Characteristic N 2 O stretch is at 2224 cm -1 , other significant peaks are due to THF that was transferred from the headspace of the reaction.  Figure S42. Vacuum transfer of the volatiles from the reaction of (DIM)Ni(NO 2 ) 2 with (Bpin) 2 bpy in a 2:3 mole ratio at 80 °C. Characteristic N 2 O stretch is at 2224 cm -1 , other significant peaks are due to THF that was transferred from the headspace of the reaction. Figure S43. 1

XPS
XPS data is summarized in Figure S44. Comparison of the bis-nitrate and bis-nitrite complexes in Fig S44b shows a 4 eV shift to lower binding energy, consistent with pentavalent nitrogen being more difficult to ionize than trivalent nitrogen. The corresponding peaks at 399.8 in both the bis-nitrate and bis-nitrite can be attributed to the di-imine ligand nitrogens. When the bisnitrate and bis-nitrite complexes are compared to the dimer, [(DIM)Ni(NO)] 2 (Fig S44a-c), there is a 0.7 eV shift of the Ni 2p binding energy to a lower binding energy, consistent with a more reduced nickel metal center in the dimer. Furthermore, the nitrosyl nitrogen is resolved from the ligand based nitrogens as a shoulder on the larger di-imine nitrogen peak. The binding energy for the nitrosyl nitrogen matches the binding energy for the nitrite nitrogens in the bis-nitrite complex, supporting a cationic nitrosyl with a nitrogen oxidation state of +3. The O1s peak of [(DIM)Ni(NO)] 2 is shifted by 0.57 eV to lower binding energy in comparison to the nitrate and nitrite starting materials. This can be rationalized by a large degree of back donation into the linear nitrosyl π * orbital, which is supported by the low NO stretching frequency. Fig S44 d-f compares the dimer to (DIM)Ni(NO)(NO 2 ), and the Ni2p region does not appear to be shifted; however, there is a diminishing of the satellite peaks. The N1s region for (DIM)Ni(NO)(NO 2 ) is asymmetric and broad, which suggests multiple types of nitrogen are present, despite lower resolution than in the other samples. The shoulder tailing towards higher binding energies indicates oxidized nitrogen, either arising from nitrite or the linear nitrosyl ligand. The O1s region for (DIM)Ni(NO)(NO 2 ) deviates greatly from symmetric, indicating incipient resolution of multiple types of oxygen. The relative intensity of the shoulder at higher binding energies is 1:2.  Typical bond lengths presented above 14 are in line with increased electron density in the α-diimine ligand for the [(DIM)Ni(NO)] 2 dimer compared to the other nickel complexes reported in the paper. All complexes other than the dimer have C-C and C=N bond lengths consistent with a neutral, unreduced ligand.

Data for (DIM)Ni(NO 3 ) 2 (MSC #19120) CCDC 2083107
General details A light green crystal (plate, approximate dimensions 0.31 × 0.19 × 0.09 mm 3 ) was placed onto the tip of a MiTeGen pin and mounted on a Bruker Venture D8 diffractometer equipped with a PhotonIII detector at 100.0 K.

Data collection
The data collection was carried out using Mo K radiation ( = 0.71073 Å, graphite monochromator) with a frame time of 0.75 seconds and a detector distance of 50 mm. A collection strategy was calculated and complete data to a resolution of 0.75 Å with a redundancy of 3 were collected. A total of 816 frames were collected. The total exposure time was 0.17 hours. The frames were integrated with the Bruker SAINT 4 software package using a narrow-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 54924 reflections to a maximum θ angle of 28.32° (0.75 Å resolution), of which 12914 were independent (average redundancy 4.253, completeness = 99.4%, R int = 17.84%, R sig = 18.52%) and 7264 (56.25%) were greater than 2σ(F 2 ). The final cell constants of a = 14.981(2) Å, b = 13.932(2) Å, c = 25.176(3) Å, β = 97.521(4)°, volume = 5209.4(12) Å 3 , are based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the Multi-Scan method (SADABS). 5 The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.7640 and 0.9220. Please refer to Table S2 for additional crystal and refinement information.

Structure solution and refinement
The space group P 1 21/c 1 was determined based on intensity statistics and systematic absences. The structure was solved using the SHELX suite of programs 6 and refined using fullmatrix least-squares on F 2 within the OLEX2 suite. 7 An intrinsic phasing solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining non-hydrogen atoms. All nonhydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final full matrix least squares refinement converged to R1 = 0.0942 and wR2 = 0.2307 (F 2 , all data). The goodness-of-fit was 1.028. On the basis of the final model, the calculated density was 1.419 g/cm 3 and F(000), 2322 e -.

Data for (DIM)Ni(NO 2 ) 2 (MSC #19121) CCDC 2083108
General details Single crystals suitable for X-ray diffraction were grown by vapor diffusion of pentane into a tetrahydrofuran solution. A light brown crystal (plate, approximate dimensions 0.14 × 0.11 × 0.07 mm 3 ) was placed onto the tip of a MiTeGen pin and mounted on a Bruker Venture D8 diffractometer equipped with a Photon III detector at 100.0 K.

Data collection
The data collection was carried out using Mo K radiation ( = 0.71073 Å, graphite monochromator) with a frame time of 5 s for low angle frames and 40 s for high angle frames. The detector distance was 130 mm. A collection strategy was calculated and complete data to a resolution of 0.75 Å with a redundancy of 7 were collected. A total of 2728 frames were collected.  Table  S3 for additional crystal and refinement information.

Structure solution and refinement
The space group Pca2 1 was determined based on intensity statistics and systematic absences. The structure was solved using the SHELX suite of programs 6 and refined using full-matrix leastsquares on F 2 within the OLEX2 suite. 7 An intrinsic phasing solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining non-hydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final full matrix least squares refinement converged to R1 = 0.0620 and wR2 = 0.1196 (F 2 , all data). The goodness-of-fit was 1.051. On the basis of the final model, the calculated density was 1.346 g/cm 3 and F(000), 1984 e -.

Data for (DIM)Ni(NO)(ONO) (MSC #20009) CCDC 2083111
General details A green crystal (approximate dimensions 0.25 × 0.12 × 0.11 mm 3 ) was placed onto the tip of a 0.05 mm diameter glass capillary and mounted on a Bruker Venture D8 diffractometer equipped with a PhotonIII detector at 100(2) K.

Data collection
The data collection was carried out using Mo Kradiation (graphite monochromator) with a frame time of 2 seconds and a detector distance of 4.00 cm. A collection strategy was calculated and complete data to a resolution of 0.71 Å with a redundancy of 3.4 were collected. Five major sections of frames were collected with 0.50º  and  scans. A total of 1640 frames were collected. The total exposure time was 1.27 hours. The frames were integrated with the Bruker SAINT software package 4 using a narrow-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 51978 reflections to a maximum θ angle of 28.32° (0.75 Å resolution), of which 5558 were independent (average redundancy 9.352, completeness = 99.8%, R int = 7.45%, R sig = 4.20%) and 4368 (78.59%) were greater than 2σ(F 2 ). The final cell constants of a = 11.6061(7) Å, b = 16.8020(11) Å, c = 12.3282(8) Å, β = 111.731(2)°, volume = 2233.2(2) Å 3 , are based upon the refinement of the XYZ-centroids of 9002 reflections above 20 σ(I) with 6.014° < 2θ < 56.34°. Data were corrected for absorption effects using the Multi-Scan method (SADABS). 5 The ratio of minimum to maximum apparent transmission was 0.900. The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.8070 and 0.9080. Please refer to Table S4 for additional crystal and refinement information.

Structure solution and refinement
The space group P2 1 /n was determined based on intensity statistics and systematic absences. The structure was solved and refined using the SHELX suite of programs. 6 An intrinsic-methods solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining nonhydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final anisotropic full-matrix least-squares refinement on F 2 with 279 variables converged at R1 = 4.30%, for the observed data and wR2 = 10.85% for all data. The goodness-of-fit was 1.114. The largest peak in the final difference electron density synthesis was 0.460 e -/Å 3 and the largest hole was -0.515 e -/Å 3 with an RMS deviation of 0.088 e -/Å 3 . On the basis of the final model, the calculated density was 1.354 g/cm 3 and F(000), 960 e -

Data collection
The data collection was carried out using Mo Kradiation (graphite monochromator) with a frame time of 40 seconds and a detector distance of 5.00 cm. A collection strategy was calculated and complete data to a resolution of 0.83 Å with a redundancy of 6 were collected. Six major sections of frames were collected with 0.50º  and  scans. A total of 1844 frames were collected. The total exposure time was 18.54 hours. The frames were integrated with the Bruker SAINT software package 4 using a narrow-frame algorithm. The integration of the data using a triclinic unit cell yielded a total of 12098 reflections to a maximum θ angle of 25.26° (0.83 Å resolution), of which 3478 were independent (average redundancy 3.478, completeness = 97.8%, R int = 20.83%, R sig = 28.27%) and 1648 (47.38%) were greater than 2σ(  Table S5 for additional crystal and refinement information.

Structure solution and refinement
The space group P-1 was determined based on intensity statistics and the lack of systematic absences. The structure was solved and refined using the SHELX suite of programs. 6 An intrinsicmethods solution was calculated, which provided most non-hydrogen atoms from the E-map. Fullmatrix least squares / difference Fourier cycles were performed, which located the remaining nonhydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final anisotropic full-matrix least-squares refinement on F 2 with 253 variables converged at R1 = 12.38%, for the observed data and wR2 = 34.84% for all data. The goodness-of-fit was 0.968. The largest peak in the final difference electron density synthesis was 2.711 e -/Å 3 and the largest hole was -1.158 e -/Å 3 with an RMS deviation of 0.160 e -/Å 3 . On the basis of the final model, the calculated density was 1.387 g/cm 3 and F(000), 434 e -.

Data collection
The data collection was carried out using Mo Mo K radiation ( = 0.71073 Å, graphite monochromator) with a frame time of 5 seconds and a detector distance of 40 mm. A collection strategy was calculated and complete data to a resolution of 0.77 Å with a redundancy of 5.3 were collected. The frames were integrated with the Bruker SAINT 4 software package using a narrowframe algorithm. The integration of the data using a monoclinic unit cell yielded a total of 92275 reflections to a maximum θ angle of 27.48° (0.77 Å resolution), of which 6025 were independent (average redundancy 15.315, completeness = 100.0%, Rint = 11.85%, Rsig = 4.31%) and 4579 (76.00%) were greater than 2σ(  Table S6 for additional crystal and refinement information.

Structure solution and refinement
The space group P 1 21/n 1 was determined based on intensity statistics and systematic absences. The structure was solved using the SHELX suite of programs 6 and refined using fullmatrix least-squares on F 2 within the OLEX2 suite. 7 An intrinsic phasing solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining non-hydrogen atoms. All nonhydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final full matrix least squares refinement converged to R1 = 0.0471 and wR2 = 0.1069 (F 2 , all data). The goodness-of-fit was 1.039. On the basis of the final model, the calculated density was 1.400 g/cm 3 and F(000), 1153 e -. The structure has disorder, with Ni coordinated to a triflate molecule as the major disorder component (0.97) and to a chlorine anion as the minor disorder component (0.03).  (2) K.

Data collection
The data collection was carried out using Mo K radiation (graphite monochromator) with a frame time of 17 seconds and a detector distance of 50 mm. A collection strategy was calculated and complete data to a resolution of 0.77 Å with a redundancy of 5.6 were collected. Four major sections of frames were collected with 0.50º  and  scans. A total of 1018 frames were collected. The total exposure time was 4.81 hours. The frames were integrated with the Bruker SAINT software package 4 using a narrow-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 21147 reflections to a maximum θ angle of 27.53° (0.77 Å resolution), of which 3630 were independent (average redundancy 5.826, completeness = 99.6%, R int = 6.87%, R sig = 5.04%) and 2816 (77.58%) were greater than 2σ(F 2 ). The final cell constants of a = 12.209(2) Å, b = 7.5518(13) Å, c = 17.864(3) Å, β = 106.243(5)°, volume = 1581.3(5) Å 3 , are based upon the refinement of the XYZ-centroids of 3676 reflections above 20 σ(I) with 4.725° < 2θ < 49.42°. Data were corrected for absorption effects using the Multi-Scan method (SADABS). 5 The ratio of minimum to maximum apparent transmission was 0.843. The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.6910 and 0.9180. Please refer to Table S7 for additional crystal and refinement information.

Structure solution and refinement
The space group P2 1 /n was determined based on intensity statistics and systematic absences. The structure was solved and refined using the SHELX suite of programs. 6 An intrinsic-methods solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining nonhydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final anisotropic full-matrix least-squares refinement on F 2 with 194 variables converged at R1 = 4.67%, for the observed data and wR2 = 11.08% for all data. The goodness-of-fit was 1.033. The largest peak in the final difference electron density synthesis was 1.351 e -/Å 3 and the largest hole was -0.663 e -/Å 3 with an RMS deviation of 0.088 e -/Å 3 . On the basis of the final model, the calculated density was 1.444 g/cm 3 and F(000), 724 e -  General details A blue crystal (approximate dimensions 0.23 × 0.18 × 0.02 mm 3 ) was placed onto the tip of a 0.05 mm diameter glass capillary and mounted on a Bruker Venture D8 diffractometer equipped with a PhotonIII detector at 123(2) K.

Data collection
The data collection was carried out using Mo Kradiation (graphite monochromator) with a frame time of 3 seconds and a detector distance of 4.00 cm. A collection strategy was calculated and complete data to a resolution of 0.71 Å with a redundancy of 11 were collected. Eight major sections of frames were collected with 0.50º  and  scans. A total of 1823 frames were collected. The total exposure time was 0.93 hours. The frames were integrated with the Bruker SAINT software package 4 using a narrow-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 48522 reflections to a maximum θ angle of 30.06° (0.71 Å resolution), of which 4277 were independent (average redundancy 11.345, completeness = 99.9%, R int = 7.82%, R sig = 3.75%) and 3269 (76.43%) were greater than 2σ(F 2 ). The final cell constants of a = 10.2092(5) Å, b = 11.9426(6) Å, c = 11.9808(5) Å, β = 91.560(2)°, volume = 1460.21(12) Å 3 , are based upon the refinement of the XYZ-centroids of 9722 reflections above 20 σ(I) with 4.816° < 2θ < 57.28°. Data were corrected for absorption effects using the Multi-Scan method (SADABS). 5 The ratio of minimum to maximum apparent transmission was 0.880. The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.7480 and 0.9740. Please refer to Table S8 for additional crystal and refinement information.

Structure solution and refinement
The space group P2 1 /c was determined based on intensity statistics and systematic absences. The structure was solved and refined using the SHELX suite of programs. 6 An intrinsic-methods solution was calculated, which provided most non-hydrogen atoms from the E-map. Full-matrix least squares / difference Fourier cycles were performed, which located the remaining nonhydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. The hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. The final anisotropic full-matrix least-squares refinement on F 2 with 176 variables converged at R1 = 3.98%, for the observed data and wR2 = 9.63% for all data. The goodness-of-fit was 1.040. The largest peak in the final difference electron density synthesis was 0.650 e -/Å 3 and the largest hole was -0.726 e -/Å 3 with an RMS deviation of 0.095 e -/Å 3 . On the basis of the final model, the calculated density was 1.409 g/cm 3 and F(000), 644 e -  Although the hyponitrite is lower in energy, our experimental results suggest that N-N bond formation does not happen between oxyanions at the same metal center.
(DIM)Ni(NO) 2 was geometry optimized as a singlet state both from two linear, and one bent and one linear starting structures and both optimized to the same structure, one bent and one nearly ( NiNO 166 o ) linear. The bent NO occupies an apical position in a trigonal pyramidal structure ∠ with the linear NO nearly in the NiDIM plane. The linear NO has an N-Ni distance shorter by 0.2 Å than to the bent NO. In sum this electronic structure is adequately described as planar (DIM)Ni(NO) radical with bent NO apical, hence 18 valence electrons with one anionic and one cationic nitrosyl. The bent/linear nature of the calculated dinitrosyl complex may help facilitate N-N bond formation, with the linear nitrosyl having electrophilic character and the bent nitrosyl having nucleophilic character.
The (DIM)Ni(NO) 2 dinitrosyl triplet is lower in energy than the singlet state by 7.3 kcal/mol. Although the singlet has one bent and one linear nitrosyl, the triplet species is pseudo-two fold symmetric, with each Ni-N-O angle ~139°, which is in line with the expectation for neutral NO. From the corresponding orbital diagram as well as the spin density plot, there is considerable amount of unpaired spin located on the nitrosyl ligands. We hypothesize that the radical character on the nitrosyl ligands could play an important role in the N-N coupling step to form N 2 O.