Hailee Cox
Christopher Dade
Jessica Metter
Steric Effects of MnDIPIMPLigand on Catalytic Reduction of CO2
The electrocatalytic reduction of carbon dioxide (CO2) to carbon monoxide (CO) is a useful pathway to capture and store atmospheric CO2. However, previous catalytic studies have found a tradeoff between reducedcatalyst dimerization via increased steric hindrance and an increase in reduction potential.[1]-Diimine ligand IP (2-(phenylimino)pyridine) derivatives, such as MnBr(CO3)(IP), provide a novel catalyst structure with the ability to optimize steric and electronic properties independently.1The sterically bulky ligand (2-[((2,6-diisopropylphenyl)imino)methyl]pyridine) (DIPIMP) bound to the Mn(I) complex demonstrates that steric and electronic effects of a catalyst can be separated and refined independently of one another, opening up a broad array of future studies into optimized CO2 reduction catalysts.1
The IP derivative with the DIPIMP ligand can be synthesized from the pro-catalyst (P) (Scheme 1). The active anion catalyst (A) can be synthesized from P, opening two reaction pathways. In inert atmospheres, a dimer (D) is formed by most IP derivatives. However, this is not a favorable pathway with the DIPIMPligand due to steric hindrance from the bulk of its side chains. When saturated CO2is present, A can react with CO2 and hydrogen atoms to produce water and CO, ultimately forming a complex that will resembleA with a positive charge. This molecule can react with various compounds to produceacetonitrile and water complexed minor products. Water from the catalytic formation of CO can react with CO2producing a bicarbonate ion (HCO3-), which canreact with debrominated P to form the bicarbonate complex.
In an inert atmosphere of Ar-saturated acetonitrile/0.2 M [Bu4N][PF6], the direct formation of the five-coordinated anionAwithout dimerization can be observed via infrared spectroelectrochemistry (IR-sec) using an optically transparent thin-layer electrochemical (OTTLE) cell. Figure 1 shows IR-sec measurements of the reduction ofP, the parent compound,toA. Initially, only peaks assigned to Pwere observed at 2030, 1950, and 1925 cm-1. After some reduction had occurred (red line), these peaks decreased in intensity,and a peak began to grow at 1823 cm-1 assigned to A. The side reaction formation of H, the water-coordinated complex, was also confirmed by the emergence of peaks at 2500 and 1960 cm-1. As the reduction proceeded, the P peaks decreased as the final anion A peak grew (green line). At this stage in the reduction of P, the second side reaction formation of the solvent coordinated radical (M) was also confirmed by the growth of peaks at 2007 and 1900 cm-1.[2] At the end of the reduction (blue line), no P peaks were observed, and only minimal M peaks were observed, indicating the successful reduction of P toA. The absence of any D peaks[3],[4] in the IR-sec spectrum supports the strong hindrance of dimerization by the DIPIMP ligand. However, the 2-electron reduction of P to A is not the only reaction taking place in situ, and the small but not insignificant complexing of debrominated P with both water and solvent, as indicated in Scheme 1, was confirmed.
While MnDIPIMP exhibited a highly desirable preference for A formation without dimerization under an inert atmosphere, it exhibited unique properties while undergoing reduction in the presence of saturated CO2. Figure 2 shows IR-sec measurements accompanying the reduction of P in situ in CO2-saturated acetonitrile/0.2 M [Bu4N] [PF6].Only P wasobserved before reduction (black line). At the start of the reduction (red line),A began to form in addition to the water-coordinated complex H, similar to the reduction of MnDIPIMP in an inert atmosphere. Further reduction (green line) led to increasing A concentration, the formation of acetonitrile-coordinated M, and the elimination of H. Towards the end of reduction (blue line), a signal at 1900 cm-1 indicated the presence of pentacarbonyl [Mn(CO)5]- (C) indirectly confirming the formation of some CO,given CO present in the cell replaces the -diimine ligand in A to form C. However, the presence of a meta-stable concentration of A, and the formation of free bicarbonate (F) and subordinate formate (S) indicated by peaks at 1680,1650, and 1610 cm-1, confirm the poor catalytic efficiency of MnDIPIMP to reduce CO2 to CO. In advanced stages of reduction (light blue line), the concentration of F and S increased significantly while the concentration of A remained relatively unchanged and the concentration of C increased minimally. At no point in the reaction were peaks characteristic of a stable bicarbonate complex[5], which can be reduced to recover the catalyst and generate CO, observed. This is further proof MnDIPIMP is too sterically hindered to react with saturated CO2 to form the bicarbonate complex from which CO and the catalyst can be recovered.
1
Hailee Cox
Christopher Dade
Jessica Metter
Appendix 1
To recreate Figure 1 from the graph in the source paper, WebPlotDigitizer was used to digitize the source lines and identify data values. An Excel sheet was then set up with the minimum number of Gaussian functions needed for each line as indicated in the assignment. Each Gaussian was given a center (max), a width (std.), and a height. These values were used in the NORM.DIST function on Excel to calculate values of each Gaussian and then normalized with respect to the height and maximum Gaussian value. A sum of normalized values (normalized sum) for each wavelength was then computed. While the assignment asked for a minimum step size x=5, it was found that a smaller step size, x=2 or x=1, produces a substantially better fit. The values of the normalized sums were compared to the digitized values at each wavelength by examining the square of their differences. This was done to ensure only one “best” solution was mathematically possible and to increase the penalty for a particularly poor estimation and fit. These values were then minimized using the Solver function on Excel to vary the max, std., and height values of each Gaussian. Additional Gaussians were added as “hidden Gaussians”—ones that influenced the shape of the line but were not individually distinguishable because they existed beneath the observable peaks. This method of iterative solutions worked quite well in producing lines that matched quite closely with the digitized data. There was some error in the fit given the sharp peaks and instrumental noise in the source graph, and even smaller step sizes may be required to capture the subtle details of the source lines.
1
Hailee Cox
Christopher Dade
Jessica Metter
Scheme 1: The stepwise formation of the five-coordinated anion (A) from the parent complex (P) coordinated to the ligand DIPIMP and its catalytic reduction of CO2 to CO. Two main side reaction pathways are important to the catalytic efficiency of A. One is the dimerization of A into dimer (D)in an inert atmosphere. In a CO2rich atmosphere, a new catalytic pathwayopens, resulting in the production of CO and H2O (top boxed reaction). H2O produced can also react with CO2 and A to form the bicarbonate complexdirectly. This positively charged precursor to A can also react with water to form H and the acetonitrile solution to form M in low concentrations.
Figure 1. IR-spectroelectrochemistry (IR-sec) spectroscopy of the reduction of MnDIPIMP in Ar-saturated acetonitrile/0.2 M [Bu4N] [PF6]. For black spectrum (λmax, m [cm-1], hm, σm [cm-1], nm):2028.059, 0.02049, 0.746, 0.532806465 (m=1);2028.341472, 0.145918633, 3.168149358, 0.125193504 (m=2);2025.848074, 0.03134089, 9.871004747,0.040410783(m=3); 1999.685354, 0.0032294745.783392809, 0.068878652 (m=4);1944.21671, 0.099610925, 6.064507485, 0.065741142(m=5); 1931.985225, 0.026038683, 25.91114955, 0.015396546 (m=6); 1919.999199, 0.010228828, 3.048539311, 0.130863416(m=7);1922.354683, 0.074004486, 7.995520152, 0.049846657 (m=8). For red spectrum: 2050.581544, 0.038031109, 2.746667979, 0.143570027(m=1);2028.010668, 0.091875462, 3.066776247, 0.130084439 (m=2);2028.030291, 0.015920611, 16.8846519, 0.023627472 (m=3);2006.684431, 0.005060394, 1.982847223, 0.198664738 (m=4); 1959.504965, 0.05494549, 10.5664561, 0.03771413(m=5); 1944.301584, 0.053298793, 5.336787275, 0.074634002(m=6); 1924.378036, 0.075913358, 9.97126644, 0.039980445(m=7); 1896.740602, 0.012312278, 10.43130735, 0.038232883(m=8);1823.262768, 0.019548839, 11.20271494, 0.035601419 (m=9).For green spectrum: 2050, 0.002972821, 2.557542915, 0.155986544(m=1); 2029.000398, 0.003909448, 1.538365247, 0.259328705(m=2); 2028.41084, 0.013464177, 2.830121881, 0.139485446(m=3); 2007.77476, 0.0225, 2.8, 0.142019136(m=4); 2027.45422, 0.008138727, 6.568932589, 0.060586662(m=5); 2000, 0.003714616, 4.272834903, 0.093367118 (m=6);1957.644771, 0.010277955, 10.62027605, 0.037543203 (m=7); 1929.123865, 0.020441636, 1.311609486, 0.302809086(m=8); 1928.713527, 0.131854496, 7.244053338, 0.055028647(m=9); 1946.685604, 0.012121722, 3.573788018, 0.111198962(m=10); 1912.562645, 0.020293052, 4.489955467, 0.088431644(m=11); 1899.185139, 0.029167155, 10.08879599, 0.039536444(m=12); 1862.315336, 0.005820365, 12.38008919, 0.032214056(m=13); 1823.238279, 0.130007939, 11.45355487, 0.03482377(m=14); 1831.909453, 0.007349319, 2.576164384, 0.15476339(m=15); 1800, 0.007282024, 17.63660468, 0.022620129(m=16).For blue spectrum:2007, 0.007762619, 2.538660112, 0.157146787(m=1); 2004.487189, 0.001977879, 6.834558335, 0.05822322(m=2); 1929.195341, 0.119930255, 4.050358335, 0.098381071(m=3); 1927.146397, 0.063898983, 9.706575131, 0.041095536(m=4); 1899.705031, 0.020497826, 11.70808486, 0.034063273(m=5);1861.833254, 0.013696731, 11.88271414, 0.033570024 (m=6); 1823.672921, 0.179409055, 11.41631703, 0.034930581(m=7); 1802.326163, 0.014156003, 15.45036407, 0.025815145(m=8).
Figure 2. IR-sec spectroscopy of the reduction of MnDIPIMP in CO2-saturated acetonitrile/0.2 M [Bu4N] [PF6].When P is reduced in a CO2 saturated acetonitrile solution, the five-coordinated anion A and pentacarbonyl [Mn(CO)5]- (C) are formed in metastable concentrations. Additionally, the concentration of free bicarbonate (F) and subordinate formate (S) increases substantially at the end of reduction.
1
[1]Spall, Steven J.P.; Keane, Theo; Tory, Joanne; Cocker, Dean C.; Adams, Harry; Fowler, Hannah; Meijer, Anthony J. H. M.; Hartl, František; Weinstein, Julia A.Manganese Tricarbonyl Complexes with Asymmetric 2-Iminopyridine Ligands: Toward Decoupling Steric and Electronic Factors in Electrocatalytic CO2 Reduction. Inorganic Chemistry, 2016, 55 (24), 12568–12582.
[2]Riplinger, C.; Sampson, M. D.; Ritzmann, A. M.; Kubiak, C. P.; Carter, E. A. Mechanistic Contrasts between Manganese and Rhenium Bipyridine Electrocatalysts for the Reduction of Carbon Dioxide. J. Am. Chem. Soc.2014, 136, 16285−16298.
[3]Zeng, Q.; Tory, J.; Hartl, F. Electrocatalytic Reduction of Carbon Dioxide with a Manganese(I) Tricarbonyl Complex Containing a Nonaromatic α-Diimine Ligand. Organometallics2014, 33, 5002− 5008.
[4]Rossenaar, B. D.; Hartl, F.; Stufkens, D. J.; Amatore, C.; Maisonhaute, E.; Verpeaux, J.-N. Electrochemical and IR/UV-Vis Spectroelectrochemical Studies of fac-[Mn(X)(CO)3(i Pr-DAB)]n (n = 0, X = Br, Me, Bz; n = +1, X = THF, MeCN, nPrCN, P(OMe)3; iPrDAB = 1,4-Diisopropyl-1,4-diaza-1,3-butadiene) at Variable Temperatures: Relation between Electrochemical and Photochemical Generation of [Mn(CO)3(α-diimine)]−. Organometallics1997, 16, 4675− 4685.
[5]Sampson, M. D.; Nguyen, A. D.; Grice, K. A.; Moore, C. E.; Rheingold, A. L.; Kubiak, C. P. Manganese Catalysts with Bulky Bipyridine Ligands for the Electrocatalytic Reduction of Carbon Dioxide: Eliminating Dimerization and Altering Catalysis. J. Am. Chem. Soc. 2014, 136, 5460−5471