Is Magnetic Bistability of Carbenes a General Phenomenon? Isolation

Article pubs.acs.org/JACS

Is Magnetic Bistability of Carbenes a General Phenomenon? Isolation of Simple Aryl(trifluoromethyl)carbenes in Both Their Singlet and Triplet States Yetsedaw A. Tsegaw,† Pritam E. Kadam,† Niklas Tötsch,‡,§ Elsa Sanchez-Garcia,*,‡,§ and Wolfram Sander*,† †

Lehrstuhl für Organische Chemie II, Ruhr-Universität Bochum, 44780 Bochum, Germany Max-Planck-Institut für Kohlenforschung, 45470 Mülheim an der Ruhr, Germany § Universität Duisburg-Essen, Universitätsstrasse 2, 45141 Essen, Germany ‡

S Supporting Information *

ABSTRACT: p-Tolyl(trifluoromethyl)carbene and the related fluorenyl(trifluoromethyl)carbene were synthesized in solid argon and characterized by IR, UV−vis, and electron paramagnetic resonance spectroscopy as well as by quantum mechanical calculations. The carbenes can be generated in both their triplet and singlet states, and both states coexist under the conditions of matrix isolation. According to our calculations, the singlet and triplet states of these carbenes are energetically nearly degenerate in the gas phase. Warming of matrices containing pure triplet p-tolyl(trifluoromethyl)carbene from 3 to 25 K leads to an interconversion of up to 20% of the triplet into the singlet state. This interconversion is thermally irreversible, and cooling back to 3 K does not change the singlet to triplet ratio. Irradiation at 365 nm results in a complete singlet to triplet interconversion, whereas 450 nm irradiation produces again up to 20% of the singlet state. An alternative way to generate the singlet carbene is the reaction of the triplet with water molecules by annealing water-doped matrices at 25 K. This results in the irreversible formation of a hydrogen-bonded complex between the singlet carbene and water. For fluorenyl(trifluoromethyl)carbene, very similar results are obtained, but the yield of the singlet state is even higher. Magnetic bistability of carbenes seems to be a general phenomenon that only depends on the singlet−triplet gap rather than on the nature of the carbene.

■

INTRODUCTION A paradigm of carbene chemistry is that carbene reactions are spin-specific.1 Since the intersystem crossing (ISC) rates between the singlet and triplet states of arylcarbenes are generally faster2 than diffusion processes in solution, the reaction pattern of these carbenes mainly depends on the population of the equilibrated spin states, and thus on the free energy singlet−triplet gap ΔGST and the temperature. At cryogenic temperatures, excited spin states are not populated, and thus, carbenes are isolated in low-temperature matrices in their ground spin states, either singlet or triplet, depending on the nature of the carbene. In contradiction to this general belief, McMahon et al. reported in 1998 that 2-naphthyl(carbomethoxy)carbene (1) can be matrix-isolated and spectroscopically characterized in both its singlet (S-1) and triplet (T-1) states.3 The singlet state S-1 was populated by visible light irradiation (λ > 515 nm) of T-1 and has a lifetime of several hours under these conditions (Scheme 1). The energy barrier preventing the expected rapid ISC back to T-1 was attributed to pronounced conformational differences between S-1 (carbomethoxy group perpendicular to the naphthyl group) and T-1 (both groups in one plane). © 2017 American Chemical Society

Scheme 1. Magnetically Bistable Carbenes 1 and 2

Toscano et al. investigated carbene 1 by time-resolved IR spectroscopy and detected both T-1 and S-1 as transient species in Freon-113.4 From these experiments, ΔGST = 0.2 ± 0.1 kcal/mol was estimated, which is much smaller than the singlet−triplet splitting ΔEST = 4.52 kcal/mol calculated by density functional theory (DFT). This discrepancy reflects both limitations of the DFT calculations and solvent effects preferentially stabilizing S-1. In the more polar acetonitrile as Received: July 2, 2017 Published: August 8, 2017 12310

DOI: 10.1021/jacs.7b06868 J. Am. Chem. Soc. 2017, 139, 12310−12316

Article

Journal of the American Chemical Society solvent, S-1 is slightly preferred.5 Later, Platz et al. studied carbene 1 using ultrafast UV−vis and IR spectroscopy.6 They found that in CHCl3, CH3CN, and Freon-113 only S-1 is present, whereas in the unpolar cyclohexane S-1 decays with a half-life of 3 ns to T-1. Thus, in both cyclohexane at room temperature and solid argon at 10 K, T-1 is the ground state, but the lifetime of S-1 increases dramatically by more than 12 orders of magnitude. Recently, we reported on bis(p-methoxyphenyl)carbene (2), a second carbene that could be isolated in rare gas matrices both in its S state and in its T state (Scheme 1).7 At temperatures below 10 K, both spin states coexist indefinitely, and irradiation at 450 nm results in an increase of S-2 and decrease of T-2, whereas 365 nm irradiation shifts the photostationary equilibrium toward T-2. Annealing at higher temperatures (>20 K) leads to an increase of S-2, which is not reversible by cooling back to very low temperatures (3 K). This indicates a thermal activation barrier for the S−T interconversion. This magnetic bistability is astonishing, since the geometries of S-2 and T-2 are very similar, differing mainly in the bond angle at the carbene center (estimated to be 119° for S-2 and 141° for T-2), and therefore, the spin interconversion does not require a large geometrical change as in 1. An open question is whether the magnetic bistability is a general phenomenon that occurs in all carbenes with small S−T gaps at cryogenic temperatures, or whether it requires specific substituents and specific electronic structures to slow the ISC. This question is of fundamental importance for the understanding of carbene reactions, but also for the development of applications such as optically switchable organic magnetic materials or tailored photoaffinity labeling. A disadvantage of carbenes 1 and 2 for studying magnetic bistability is that both carbenes are rather flexible and form several energetically close lying conformers. In addition, 1 shows secondary photochemistry, which results in complex product mixtures that complicate the interpretation of the spectra. We therefore were searching for conformationally simple, photostable arylcarbenes with nearly degenerate S and T states that would allow us to matrix-isolate both spin states. Phenylcarbene (3) is highly photolabile toward rearrangement to 1,2,4,6-cycloheptatetraene,8 but phenyl(trifluoromethyl)carbene (4a) is photochemically rather stable toward visible light irradiation under the conditions of matrix isolation.9 Carbene 4a rapidly reacts with molecular oxygen, as expected for a triplet-ground-state carbene.10,11 The triplet ground state was further confirmed by electron paramagnetic resonance (EPR) spectroscopy.9 According to calculations by Song and Sheridan, the trifluoromethyl group stabilizes the singlet state by a small amount (ΔEST(3) = 5.4 kcal/mol, ΔEST(4a) = 4.0 kcal/mol).12 Aryl(trifluoromethyl)carbenes 4 are of special interest since they found wide applications in photoaffinity labeling.13 The photochemistry and the excited-state dynamics of these carbenes were extensively studied by time-resolved spectroscopy.14−19 The S−T gap in 4a is still too large to expect magnetic bistability, but it can be further lowered by electron-donating substituents in the para-position of the phenyl ring. Here, ptolyl(trifluoromethyl)carbene (4b) was selected as a feasible candidate for a low-gap carbene. Platz et al. reported the synthesis of carbene 4b by photolysis of diazirine 8 in a perfluorated organic glass at 77 K (Scheme 2).14 Under these conditions, a typical triplet carbene EPR spectrum with zero

field splitting (zfs) parameters D = 0.5215 cm−1 and E = 0.0249 cm−1 was observed, which suggests that 4b has either a triplet ground state or such a small S−T gap that the excited triplet state is thermally populated even at 77 K. From laser flash photolysis studies of 8 in the presence of various quenchers, it was concluded that at room temperature the singlet state S-4b and triplet state T-4b are in a rapid thermal equilibrium, and thus, 4b can be trapped with both typical singlet and triplet quenchers. From these experiments, the S−T gap of 4b was estimated to be ΔGST = 0.5−1.5 kcal/mol.14 The related fluorenyl(trifluoromethyl)carbene (4c) was studied by Platz et al. using ultrafast time-resolved absorption spectroscopy.16 In acetonitrile, the singlet carbene S-4c was characterized as a transient species with an initial absorption maximum at 427 nm that shifts to 432 nm after vibrational cooling. Scheme 2. Synthesis and Chemistry of Carbene 4b

Carbenes 4b and 4c are simple arylcarbenes that should allow us to answer the question of whether magnetic bistability is a general phenomenon in carbenes with small S−T gaps. We therefore studied these carbenes under the conditions of matrix isolation. The principal results are identical for both carbenes, and therefore, we here discuss only 4b in detail; for 4c, we refer to the Supporting Information.

■

RESULTS AND DISCUSSION Carbene 4b was generated by either UV (365 nm) photolysis of p-tolyl(trifluoromethyl)diazirine (8) or visible light (450 nm) photolysis of p-tolyl(trifluoromethyl)diazomethane (9) in neon, argon, xenon, or nitrogen matrices at 3−4 K. The carbene 4b was characterized by IR, UV−vis, and EPR spectroscopy. Photolysis of 8 produced small amounts of 9; 12311


Article

Journal of the American Chemical Society

characteristic absorptions at 1578.8 and 1599.5 cm−1 , respectively (Figure 2). Annealing at 25 K or prolonged 450

otherwise, the same products were obtained from both precursors (Scheme 2). EPR Experiments. Photolysis of 9 in argon at 4 K results in the formation of T-4b with an EPR spectrum typical of a triplet carbene (Figure 1). The zfs parameters |D/hc| = 0.5219 cm−1

Figure 1. X-band continuous wave (CW) EPR spectra of an argon matrix showing the thermal behavior of T-4b after warming from 4 to 25 K and cooling back to 4 K. (a) Argon matrix at 4 K showing the spectrum of T-4b (red line; zfs parameters |D/hc| = 0.5219 cm−1 and | E/hc| = 0.0304 cm−1). (b) After subsequent annealing at 25 K for 10 min and cooling back to 4 K, 36% of the signal intensity is lost (blue line). The inset shows the loss of the T-4b signal at the z1 canonical position after cooling back from 8, 12, 16, 20, and 25 to 4 K. The bands marked with an asterisk are assigned to products of carbene− carbene rearrangements. Figure 2. IR spectra showing the interconversion of S-4b and T-4b in argon matrices. (a) IR spectrum after several hours of 450 nm irradiation of 9 at 3 K (dark green). (b) Difference IR spectrum at 3 K showing changes after 60 min of 365 nm irradiation of the same matrix. Bands pointing downward, assigned to S-4b, are disappearing, and bands pointing upward, assigned to T-4b, are appearing. (c) Difference IR spectrum of the same matrix after 10 min of annealing to 30 K and cooling back to 3 K. Bands pointing downward, assigned to T-4b, are disappearing, and bands pointing upward, assigned to S-4b, are appearing. (d) Difference IR spectrum at 3 K showing changes after 60 min of 365 nm irradiation of the same matrix. Bands pointing downward, assigned to S-4b, are disappearing, and bands pointing upward, assigned to T-4b, are appearing. (e) Difference IR spectrum at 3 K showing changes after 60 min of 450 nm irradiation of the same matrix. Bands pointing downward, assigned to T-4b, are disappearing, and bands pointing upward, assigned to S-4b, are appearing. (f) Superimposed IR spectrum showing the interconversion of T-4b (blue) to S-4b (red) calculated at the B2PLYP-GD3BJ/aug-cc-pVTZ level of theory. Equal contributions of S-4b and T-4b were used to simulate the IR spectra.

and |E/hc| = 0.0304 cm−1 are similar to those reported for T-4b in an organic glass at 77 K.14 The D values are almost identical, whereas the E value in argon is slightly larger than that recorded in the organic glass, which might reflect small differences in the conformation of T-4b generated in different matrices. In an attempt to measure the Curie−Weiss behavior and thus to determine the ground state of 4b, the matrix was stepwise annealed for 10 min at temperatures of 8, 12, 16, 20, and 25 K. An EPR spectrum was recorded at each of these temperatures, and before the temperature was increased to the next higher level, the matrix was cooled back to 4 K to record a second EPR spectrum (Figure 1, inset). These experiments clearly demonstrate an irreversible loss of T-4b at each temperature step, summing to more than 30% loss after warming to 25 K. While keeping the matrix at 4 K for several hours did not result in any increase of the signals of T-4b, several minutes of UV irradiation (365 nm) resulted in a partial signal recovery. In these experiments, we took care that 9 was completely photolyzed, and therefore, remaining 9 was not the precursor of T-4b under these conditions. Subsequent annealing again irreversibly reduced the amount of T-4b, and this behavior was reproducible in repeated annealing/ irradiation cycles. These experiments indicate that T-4b thermally rearranges to an EPR-silent product, which upon UV irradiation rearranges back to T-4b. It is tempting to assign this EPR-silent product to the singlet state S-4b. IR Experiments. This assumption was confirmed by following the thermal and photochemical S-T interconversion of 4b by IR spectroscopy. Visible light irradiation (450 nm) of matrix-isolated 9 rapidly results in the complete photolysis of the diazo precursor and formation of a major product assigned to T-4b and a minor product assigned to S-4b with

nm irradiation leads to an increase of the bands of the minor component S-4b. UV irradiation (365 nm) results in the almost complete bleaching of all IR bands of S-4b and concurrently in an increase of the absorptions of the major product T-4b. This assignment was confirmed by comparing the experimental IR spectra to calculations at the B2PLYP-GD3BJ/aug-cc-pVTZ (Figure 2 and Figure S2 and Tables S1 and S2, Supporting Information) and B3LYP-D3/Def2-TZVP (Figure S6, Supporting Information) levels of theory, which nicely reproduce the experimental data. The highest yield of S-4b of approximately 20% was obtained in a nitrogen matrix, whereas in solid argon only 13% was obtained after 450 nm irradiation and annealing at 25 K. Further 450 nm irradiation or annealing did not increase the 12312


Article


λmax = 436 nm, and irradiation into this band results in the rearrangement of T-4b to S-4b. To confirm the assignment of the UV−vis absorptions of the two states of 4b, we recorded UV−vis and IR spectra from the same matrix (Figure 3 and Figure S4, Supporting Information). These experiments allow us to correlate changes in band intensities in the UV−vis and IR regions and confirm the band assignments. Reaction of Carbene 4b with Water. Previously, we demonstrated that the spin state of triplet-ground-state carbenes such as diphenylcarbene (5)20,21 or fluorenylidene (6)22 is switched from triplet to singlet upon formation of hydrogen bonds with methanol or water under the conditions of matrix isolation. In a similar experiment, 9 was photolyzed (450 nm, 3 K) in argon matrices doped with 0.5−1% water. Similar to the photolysis in argon in the absence of water, mixtures with T-4b as the major constituent and S-4b as the minor constituent were formed. Annealing at 20−25 K for several minutes allowed for the diffusion of water molecules,20 which can be easily followed in the IR spectra by observing the formation of water dimers and oligomers from the water monomer. Simultaneously, all IR bands assigned to T-4b decrease in intensity, and singlet carbene S-4b is formed. However, in comparison to S-4b obtained in the absence of water, all IR bands of the singlet carbene are blue-shifted by 3− 9 cm−1. This is a clear indication that the water complex S-4b··· H−OH and not the free singlet carbene is the preferred product under these conditions. A further proof for the water complex S-4b···H−OH is its 450 nm photochemistry, which rapidly leads to the insertion product p-tolyl(trifluoromethyl)ethanol (11) (Figure 4, Table 1). In contrast, as described

yields of S-4b. The interconversion between T-4b and S-4b was reproducible in several 450 nm/365 nm irradiation cycles. At 3 K and in the absence of irradiation, T-4b and S-4b coexist without noticeable changes in their spectra even after several days. A different photochemistry is observed if the S/T mixture of 4b is irradiated with the intense, highly monochromatic light of an optical parametric oscillator (OPO) laser system in the range between 435 and 444 nm (Scheme 2). Under these conditions, both the singlet and triplet states of carbene 4b decrease and 1-(trifluoromethyl)-5-methyl-1,2,4,6-cycloheptatetraene (10) is formed in high yields (Figure S7, Supporting Information). Short-wavelength UV irradiation (254 nm) leads 10 to partially rearrange back to 4b. Interestingly, even prolonged irradiation of 4b with the low-intensity light-emitting diode (LED) light source with a maximum emission at 450 nm does not result in the rearrangement to 10. UV−Vis Experiments. The singlet and triplet states of 4b are easily distinguished in their UV−vis spectra. Visible light photolysis (450 nm) of 9 produces the UV−vis spectrum shown in Figure 3 with a series of absorptions stretching from 250 to 450 nm. A broad band with λmax = 326 nm completely disappears after 365 nm irradiation and reappears after 450 nm irradiation, and thus is assigned to S-4b. The triplet carbene T4b shows an absorption with a pronounced progression with

Figure 4. IR spectra of the water complex S-4b···H−OH matrixisolated in argon. (a) Difference IR spectrum showing changes of an argon matrix (without water) containing T-4b and S-4b at 3 K after annealing for 10 min at 25 K (black line). Bands pointing downward, assigned to T-4b, are disappearing, and bands pointing upward, assigned to S-4b, are appearing. Intensity divided by 5. (b) Difference IR spectrum of a similar experiment in argon doped with 1% water (blue line). Bands pointing downward are disappearing and assigned to T-4b, and bands pointing upward are appearing and assigned to S4b···H−OH. (c) Difference IR spectrum of the same matrix showing changes after 3 h of irradiation (λ = 450) (blue line). Bands pointing downward are assigned to S-4b···HOH, and bands pointing upward are assigned to p-tolyl(trifluoromethyl)ethanol (11). (d) Reference spectrum of 11, matrix-isolated in argon at 3 K.

Figure 3. UV−vis (top) and IR (bottom) spectra recorded from the same nitrogen matrix showing the S−T interconversion of 4b. (a) Spectra of T-4b and S-4b obtained after several hours of 450 nm irradiation of 9 at 3 K (blue line). (b) Spectra obtained after 60 min of irradiation with λ = 365 nm at 3 K (red line). (c) Spectra obtained after subsequent annealing for 30 min at 20 K and cooling back to 3 K (gray line). The weak band of T-4b in the visible region of the UV−vis spectrum is shown in the inset. 12313


Article

Journal of the American Chemical Society Table 1. Experimental and Calculated Vibrational Frequencies and Shifts of the S-4b···H2O Complex calculateda S-4b ν/cm 1071.6 1137.4 1191.5 1225.3 1642.1

−1

(Iabs)

(278.7) (139.3) (645.0) (115.5) (247.8)

experimental

S-4b···HOH ν/cm−1 (Iabs) 1070.8 1131.2 1192.8 1230.5 1648.9

(278.6) (191.3) (656.8) (49.4) (290.3)

shiftb

argonc ν/cm−1 (Irel)

Ar/H2Od ν/cm−1 (Irel)

shiftb

assignmente

−0.8 −6.2 +1.3 +5.2 +6.8

1086.7(32) 1139.2 (39) 1171.7(100) 1206.4(44) 1599.5 (43)

1095.3(40) 1147.5 (50) 1178.6(100) 1211.1(86) 1602.2 (75)

+8.6 +8.3 +6.9 +4.7 +2.7

C−CF3 bend (out of plane) C−CF3 bend (in plane) C−C−C stretch (carbene center) C−H deformation (in plane) CC stretch (ring)

a

Calculated at the B2PLYP-GD3BJ/aug-cc-pVTZ (S-4b) or B2PLYP-GD3BJ/aug-cc-pVTZ//B3LYP/aug-cc-pVTZ (S-4b···HOH) level of theory. Frequency shift relative to the monomers (cm−1). cIn an argon matrix at 3 K. dIn an argon matrix doped with 1% H2O at 3 K. eTentative assignment. b

above, 450 nm irradiation of mixtures of T-4b and S-4b leads to an increase of S-4b and not to the destruction of the singlet carbene. At temperatures below 20 K, the water complex S4b···H−OH is thermally stable and does not show OH insertion via quantum chemical tunneling as in the water complexes of diphenylcarbene (5)20 and fluorenylidene (6).22 This is presumably due to the higher barrier for the OH insertion of the water complex of 4b compared to that of 5 and 6 (see below). Computational Studies on Carbene 4b. There are only a few accurate thermochemical gas-phase data available for carbenes.23−28 CCSD(T) calculations with large basis sets are often used as the “gold standard” for calculating ΔEST of carbenes29 and for calibrating less expensive DFT results. Here, our calculations of the singlet−triplet splitting of 4b at the CCSD(T)/cc-pV(D/T)Z//B97D-GD3/def2-TZVP level of theory result in ΔEST = 0.0 kcal/mol, thus predicting degeneracy of the spin states in the gas phase. We tested several density functionals (see the Supporting Information for details) and found that B2PLYP-GD3BJ predicts ΔEST = −0.4 kcal/mol (the zero point vibrational energy (ZPVE) correction is 0.03 kcal/mol at this level of theory), in excellent agreement with the CCSD(T) result. Thus, unless otherwise specified, we are using this functional for discussing the structures, energies, and spectra of 4b. While the adiabatic singlet−triplet gap is very small, the vertical gaps of S-4b and T-4b are −8.7 and 8.5 kcal/mol at their respective minimum geometries. The relaxed onedimensional scan of S-4b and T-4b reveals an intersection at a carbene C−C−C bond angle of 127°, the triplet minimum at 135°, and the singlet minimum at 115° (Figure 5). Apart from the carbene C−C−C angle, the geometries of S-4b and T-4b differ in the rotation of the methyl group (Table S7, Supporting Information). The other degrees of freedom are very similar in both states. Unlike S-4b and T-4b, the complexes S-4b···HOH and T-4b···HOH differ geometrically because the carbene− water interaction is dependent on the electronic state. As expected, interaction with a water molecule results in a large stabilization of the singlet state (10.1 kcal/mol without and 8.2 kcal/mol including ZPVE correction), whereas the triplet state is only weakly stabilized (2.9 kcal/mol without and 2.0 kcal/ mol including ZPVE correction). The calculations also allow us to rationalize the higher thermal stability of the water complex of the singlet state of 4b compared to that of diphenylcarbene (5) and fluorenylidene (6). While the barriers for the OH insertion of S-5···HOH and S-6···HOH are reported as 6.6 and 5.1 kcal/mol, respectively, at the B3LYP-D3/6-311++G(d,p) level of theory,22,30 that of S4b···HOH is predicted to 9.2 kcal/mol at the same level and 10.0 kcal/mol at the CCSD(T)/cc-pV(D/T)Z//B97D-GD3/

Figure 5. Relaxed one-dimensional scan of the C−C−C carbene angles of 4b and the water complexes 4b···H−OH calculated at the B3LYP-GD3BJ/aug-cc-pVTZ level of theory. Adiabatic energy differences are empirically corrected using the B2PLYP-GD3BJ results (see the Supporting Information for details). Solid lines represent 4b, dotted lines 4b···HOH. The irregularity at 132° in the scan of S-4b··· HOH is caused by the rotation of the CF3 group. Please note that vertical excitations cannot be extrapolated from this figure.

def2-TZVP level of theory (Figure S13, Supporting Information), and thus considerably higher. We assume that the higher activation barriers result in tunneling rates that are too slow to be observable in our experiments.

■

CONCLUSION Carbene 4b is the structurally simplest carbene that has been demonstrated to show magnetic bistability under the conditions of matrix isolation. According to the DFT calculations, the major difference between the singlet and triplet states of carbene 4b is the C−C−C bond angle at the carbene center, 12314


Article


ISC rates must be considered. Annealing at higher temperatures results in soft matrices which enable conformational relaxation of the carbenes and allow small intrinsic barriers to be overcome. We assume that this results in equilibrium populations of the singlet and triplet states. Irradiation at low temperature (3 K), on the other hand, results in nonequilibrium populations of the spin states. Thus, the combination of thermo- and photochemistry allows the spin states of the carbenes to be shifted in a controlled way. We are currently using this phenomenon to study the spin-specific intermolecular reactivity of carbenes. In summary, magnetic bistability of carbenes is a general phenomenon that only depends on the S−T gap of the carbene. This opens the way for the development of new optically switchable magnetic materials.

which increases from 115° in S-4b to 135° in T-4b. The rotation of the methyl group differs for S-4b and T-4b (Table S7); otherwise, the structures are very similar. Despite this geometric similarity, both states indefinitely coexist in cryogenic matrices and can be interconverted by irradiation into the absorption maxima of S-4b in the UV range or T-4b in the visible range, respectively. Under all experimental conditions, T-4b is the major constituent. S-4b is never obtained in more than 20% yield (thermally or upon 450 nm irradiation), whereas 365 nm irradiation transforms S-4b quantitatively to T4b. Annealing of matrices containing T-4b results in the thermal formation of S-4b; however, the reversed thermal interconversion of S-4b into T-4b is not observed, similar to S2, which thermally does not rearrange to T-2.7 In contrast, S-1 is only metastable and thermally completely rearranges to T-1.3 Since the triplet state is 3-fold degenerate, we expect a population of 75% triplet state and 25% singlet state if the singlet−triplet gap ΔEST = 0. From the observed maximum yield of 20% S-4b, we estimate ΔEST = 0.014 kcal/mol under the assumption of a Boltzmann distribution between the spin states. The Supporting Information contains more details on how we calculated ΔEST from the populations. In contrast, much higher yields of S-4c are obtained on annealing of matrices containing S/T mixtures of 4c (see Figures S8−S10, Supporting Information). This hints to a singlet ground state of 4c and a kinetically stabilized excited triplet state. The singlet−triplet splitting of 4b estimated from our experiments is in good agreement with the computational results. However, the prediction of such small energy differences is clearly beyond the accuracy of the calculations. Moreover, rare gas matrices will interact with the trapped carbenes, and the highly polar singlet states presumably will be more stabilized than the less polar triplet states. Since the stabilization of the carbenes depends on the matrix environment (different matrix sites), we expect a distribution of singlet−triplet splitting values on the order of 1−2 kcal/mol rather than a well-defined value as in the gas phase, as reported in our previous studies of diphenylcarbene (5) in argon matrices.30 Thus, matrix site effects might exceed the intrinsic energy differences between the singlet and triplet states of the trapped carbenes, resulting in the coexistence of carbenes in both spin states. Matrix effects change not only the thermodynamic stability of trapped species, but also that of transition states and therefore strongly influence the kinetics. This results in a distribution of rate constants and in the dispersive kinetics observed in our experiments (see Table S4, Supporting Information). In addition, at the very low temperatures used in our experiments (3 K), rare gas matrices are very rigid, and conformational changes as required for the singlet−triplet interconversion of carbenes are kinetically hindered. To be efficient, spin interconversion must be coupled with changes of the C−C− C bond angle, and this process is hindered in very rigid matrices at low temperatures. Again, transition-state stabilization and steric hindrance depend on the local matrix environment, resulting in a distribution of rates for the ISC as observed in our experiments. From our experiments, we conclude that carbenes with small S−T gaps can coexist in both their singlet and triplet states under the conditions of matrix isolation. The matrix plays an important role, and magnetic bistability results from both thermodynamic and kinetic effects of the matrix on the trapped carbene. In addition, small intrinsic barriers that influence the

■

EXPERIMENTAL SECTION

■

ASSOCIATED CONTENT

General Procedures. All chemicals required to prepare the precursors were commercially available. For materials and synthesis details of the precursors, see the Supporting Information. Matrix Isolation Technique. Matrix isolation measurements were performed by standard techniques as was described in preceding work.7,20 For the matrix isolation experiments, argon, neon, xenon, and nitrogen (Messer-Griesheim, 99.9999%) were exploited. The required precursor was sublimed (8 at −30 °C, 9 at 35 °C, and 11 at 0 °C) and co-deposited on the cold spectroscopic window (3 K for IR and UV− vis and 4 K for EPR) with a large excess of inert gases with a flow rate of approximately 1.90 sccm. Typically, the deposition takes approximately 60 min. The experiments in 1% water-doped argon matrices were carried out as described previously.22,30 Ultrapure water used in these experiments was degassed by several freeze−thaw cycles. Computational Details. Unless mentioned otherwise, all DFT calculations were performed with the aug-cc-pVTZ basis set31 and the Gaussian 09 package,32 and Grimme’s dispersion corrections with Becke−Johnson damping (GD3BJ)33,34 were used. Relaxed scans were carried out by fixing the carbene angle in 3° intervals. CCSD(T) single-point calculations were performed using the MOLPRO program.35 The basis set limit was approximated by extrapolating the correlation energy using cc-pVDZ and cc-pVTZ36 calculations yielding cc-pV(D/T)Z. The structures were optimized at the B97DGD3/def2-TZVP level of theory33,37 using the TURBOMOLE code (version 6.4).38,39 For further computational details, see the Supporting Information.

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b06868.

■

Experimental and computational details, supplementary figures and tables, and Cartesian coordinates (PDF)

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Yetsedaw A. Tsegaw: 0000-0002-8598-8972 Elsa Sanchez-Garcia: 0000-0002-9211-5803 Wolfram Sander: 0000-0002-1640-7505 Notes

The authors declare no competing financial interest. 12315


Article


■

(31) Kendall, R. A.; Dunning, T. H.; Harrison, R. J. J. Chem. Phys. 1992, 96, 6796−6806. (32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J., J, A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision D.01; Gaussian Inc.: Wallingford, CT, 2009. (33) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. (34) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011, 32, 1456−1465. (35) Werner, H.-J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schütz, M. Wiley Interdisciplinary Reviews: Computational Molecular Science 2012, 2, 242−253. (36) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007−1023. (37) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297−3305. (38) TURBOMOLE V6.4 2012, a development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989−2007, TURBOMOLE GmbH, since 2007. http://www.turbomole.com, 2012. (39) Ahlrichs, R.; Baer, M.; Haeser, M.; Horn, H.; Koelmel, C. Chem. Phys. Lett. 1989, 162, 165−169.

ACKNOWLEDGMENTS This work was supported by the Cluster of Excellence RESOLV (EXC 1069) funded by the Deutsche Forschungsgemeinschaft (DFG). E.S.-G. also acknowledges a Plus3 grant by the Boehringer-Ingelheim Foundation.

■

REFERENCES

(1) Eisenthal, K. B.; Moss, R. A.; Turro, N. J. Science 1984, 225, 1439−45. (2) Eisenthal, K. B.; Turro, N. J.; Aikawa, M.; Butcher, J. A., Jr.; DuPuy, C.; Hefferon, G.; Hetherington, W.; Korenowski, G. M.; McAuliffe, M. J. J. Am. Chem. Soc. 1980, 102, 6563−5. (3) Zhu, Z. D.; Bally, T.; Stracener, L. L.; McMahon, R. J. J. Am. Chem. Soc. 1999, 121, 2863−2874. (4) Wang, Y. H.; Yuzawa, T.; Hamaguchi, H. O.; Toscano, J. P. J. Am. Chem. Soc. 1999, 121, 2875−2882. (5) Wang, Y.; Hadad, C. M.; Toscano, J. P. J. Am. Chem. Soc. 2002, 124, 1761−7. (6) Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008, 112, 11093−8. (7) Costa, P.; Lohmiller, T.; Trosien, I.; Savitsky, A.; Lubitz, W.; Fernandez-Oliva, M.; Sanchez-Garcia, E.; Sander, W. J. Am. Chem. Soc. 2016, 138, 1622−1629. (8) West, P. R.; Chapman, O. L.; LeRoux, J. P. J. Am. Chem. Soc. 1982, 104, 1779−82. (9) Blanch, R. J.; Wentrup, C. Aust. J. Chem. 2015, 68, 36−43. (10) Sander, W. W. J. Org. Chem. 1988, 53, 121−126. (11) Mal’tsev, A. K.; Zuev, P. S.; Nefedov, O. M. Izv. Akad. Nauk SSSR, Ser. Khim. 1987, 463−4. (12) Song, M.-G.; Sheridan, R. S. J. Phys. Org. Chem. 2011, 24, 889− 893. (13) Tschirret-Guth, R. A.; Medzihradszky, K. F.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1999, 121, 4731−4737. (14) Admasu, A.; Gudmundsdottir, A. D.; Platz, M. S.; Watt, D. S.; Kwiatkowski, S.; Crocker, P. J. J. Chem. Soc., Perkin Trans. 2 1998, 1093−1099. (15) Noller, B.; Poisson, L.; Maksimenka, R.; Fischer, I.; Mestdagh, J.-M. J. Am. Chem. Soc. 2008, 130, 14908−14909. (16) Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008, 130, 2304−2313. (17) Wang, J.; Kubicki, J.; Peng, H.; Platz, M. S. J. Am. Chem. Soc. 2008, 130, 6604−6609. (18) Noller, B.; Hemberger, P.; Fischer, I.; Alcaraz, C.; Garcia, G. A.; Soldi-Lose, H. Phys. Chem. Chem. Phys. 2009, 11, 5384−5391. (19) Noller, B.; Poisson, L.; Maksimenka, R.; Gobert, O.; Fischer, I.; Mestdagh, J. M. J. Phys. Chem. A 2009, 113, 3041−3050. (20) Costa, P.; Sander, W. Angew. Chem., Int. Ed. 2014, 53, 5122− 5125. (21) Knorr, J.; Sokkar, P.; Schott, S.; Costa, P.; Thiel, W.; Sander, W.; Sanchez-Garcia, E.; Nuernberger, P. Nat. Commun. 2016, 7, 12968. (22) Costa, P.; Trosien, I.; Fernandez-Oliva, M.; Sanchez-Garcia, E.; Sander, W. Angew. Chem., Int. Ed. 2015, 54, 2656−2660. (23) Paulino, J. A.; Squires, R. R. J. Am. Chem. Soc. 1991, 113, 5573− 5580. (24) Lengel, R. K.; Zare, R. N. J. Am. Chem. Soc. 1978, 100, 7495− 7499. (25) Leopold, D. G.; Murray, K. K.; Miller, A. E. S.; Lineberger, W. C. J. Chem. Phys. 1985, 83, 4849−4865. (26) Poutsma, J. C.; Paulino, J. A.; Squires, R. R. J. Phys. Chem. A 1997, 101, 5327−5336. (27) Poutsma, J. C.; Nash, J. J.; Paulino, J. A.; Squires, R. R. J. Am. Chem. Soc. 1997, 119, 4686−4697. (28) Poutsma, J. C.; Upshaw, S. D.; Squires, R. R.; Wenthold, P. G. J. Phys. Chem. A 2002, 106, 1067−1073. (29) Hu, C.-H. Chem. Phys. Lett. 1999, 309, 81−89. (30) Costa, P.; Fernandez-Oliva, M.; Sanchez-Garcia, E.; Sander, W. J. Am. Chem. Soc. 2014, 136, 15625−30. 12316


Is Magnetic Bistability of Carbenes a General Phenomenon? Isolation

Recommend Documents