Error-Controlled Exploration of Chemical Reaction Networks with Gaussian Processes

Gregor N. Simm and Markus Reiher (2018)

Highlighted by Jan Jensen

See if you recognise this scenario: you benchmark a cheap method against an accurate, but expensive, method for a set of molecules and get a mean error that you can use to correct the results obtained with the cheap method. But sooner or later you  start using the cheap method on molecules that look increasingly different from your benchmark set. At what point should you do another benchmark calculation against your expensive method? Simm and Reiher use Gaussian Processes (GP) to provide a quantitative answer.

GP is a way to fit a numerical function to a set of data points with uncertainties of the fit for every point in the fit. Simm and Reiher's basic idea is to arrange the benchmark points on the x-axis by computing a distance between pairs of molecular structures and performing a GP fit. Now compute the x-coordinate of the molecule you're uncertain about: if the uncertainty in the GP fit for that point is larger than the standard deviation of the mean error of your cheap method computed for the benchmark set, then you need to benchmark your cheap method against the more expensive method for that molecule.

This work is licensed under a Creative Commons Attribution 4.0 International License.

Spectroscopic Evidence for Aminomethylene (H−C̈−NH2)—The Simplest Amino Carbene

Eckhardt, A. K.; Schreiner, P. R., Angew. Chem. Int. Ed. 2018, 57, 5248-5252
Contributed by Steven Bacharach
Reposted from Computational Organic Chemistry with permission

Eckhardt and Schreiner have spectroscopically characterized the aminomethylene carbene 1.¹ Their characterization rests on IR spectra, with comparison to the computed AE-CCSD(T)/cc-pCVQZ anharmonic vibrational frequencies, and the UV-Vis spectra, with comparison to the computed B3LYP/6–311++G(2d,2p) transitions.

1 can be converted to 2 by photolysis. Interestingly, 1 does not convert to 2 after 5 days on the matrix in the dark. This is in distinct contrast to hydroxycarbene and related other carbene which undergo quantum mechanical tunneling (see this post and this post). Examination of the potential energy surface for the reaction of 1 to 2 at AE-CCSD(T)/cc-pCVQZ (see Figure 1) identifies that the lowest barrier is 45.8 kcal mol^-1, about 15 kcal mol^-1 larger than the barrier for the hydroxycarbene rearrangement. Additionally, the barrier width for 1 → 2 is 25% larger than for the hydroxycarbenes. Both of these suggest substantially reduced tunneling, and WKB analysis predicts a tunneling half-life of more than a billion years. The stability of 1 is attributed to the strong π-donor ability of nitrogen to the electron-poor carbene. This is reflected in a very short C-N bond (1.27 Å).

Figure 1. Structures and energies of 1 and 2 and the transition states that connect them. The relative energies (kcal mol^-1) are computed at AE-CCSD(T)/cc-pCVQZ.

References

1) Eckhardt, A. K.; Schreiner, P. R., "Spectroscopic Evidence for Aminomethylene (H−C̈−NH2)—The
Simplest Amino Carbene." Angew. Chem. Int. Ed. 2018, 57, 5248-5252, DOI: 10.1002/anie.201800679.

InChIs

1: InChI=1S/CH3N/c1-2/h1H,2H2
InChIKey=KASBEVXLSPWGFS-UHFFFAOYSA-N

2: InChI=1S/CH3N/c1-2/h2H,1H2
InChIKey=WDWDWGRYHDPSDS-UHFFFAOYSA-N

'
This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.

Longest C–C Single Bond among Neutral Hydrocarbons with a Bond Length beyond 1.8 Å

Ishigaki, Y.; Shimajiri, T.; Takeda, T.; Katoono, R.; Suzuki, T., Chem 2018, 4, 795-806
Contributed by Steven Bacharach

Reposted from Computational Organic Chemistry with permission

This is a third post in a series dealing with very short or very long distances between atoms. Ishigaki, Shimajiri, Takeda, Katoono, and Suzuki have prepared three related analogues of hexaphenylethane that all have long C-C bonds.¹ The idea is to create a core by fusing two adjacent phenyls into a naphylene, and then protect the long C-C bond through a shell made up of large aryl groups, 1. Fusing another 5-member ring opposite to the stretched C-C bond (2) creates a scissor effect that should stretch that bond further, even more so in the unsaturated version 3.

Their M062-x/6-31G* computations predict an increasing longer C-C bond (highlighted in blue in the above drawing): 1.730 Å in 1, 1.767 Å in 2, and 1.771 Å in 3. The structure of 3 is shown in Figure 3.

Figure 1. M06-2x/6-31G(d) optimized structure of 3.

These three compounds were synthesized, and characterized by IR and Raman spectroscopy. Their x-ray crystal structures at 200 K and 400K were also determined. The C-C distances are 1.742 Å (1), 1.773 Å (2) and 1.798 Å (3) with distances slightly longer at 400 K. These rank as the longest C-C bonds recorded.

References

1) Ishigaki, Y.; Shimajiri, T.; Takeda, T.; Katoono, R.; Suzuki, T., "Longest C–C Single Bond among Neutral Hydrocarbons with a Bond Length beyond 1.8 Å." Chem 2018, 4, 795-806, DOI: 10.1016/j.chempr.2018.01.011.

InChIs

1: InChI=1S/C40H26/c1-5-17-32-27(11-1)23-24-28-12-2-6-18-33(28)39(32)36-21-9-15-31-16-10-22-37(38(31)36)40(39)34-19-7-3-13-29(34)25-26-30-14-4-8-20-35(30)40/h1-26H
InChIKey=IFIFQLGAOZULMX-UHFFFAOYSA-N

2: InChI=1S/C42H28/c1-5-13-33-27(9-1)17-18-28-10-2-6-14-34(28)41(33)37-25-23-31-21-22-32-24-26-38(40(37)39(31)32)42(41)35-15-7-3-11-29(35)19-20-30-12-4-8-16-36(30)42/h1-20,23-26H,21-22H2
InChIKey=HSJDZFLRPWJAGF-UHFFFAOYSA-N

3: InChI=1S/C42H26/c1-5-13-33-27(9-1)17-18-28-10-2-6-14-34(28)41(33)37-25-23-31-21-22-32-24-26-38(40(37)39(31)32)42(41)35-15-7-3-11-29(35)19-20-30-12-4-8-16-36(30)42/h1-26H
InChIKey=QQYNKBIOZSXWGD-UHFFFAOYSA-N

'
This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.