James E. "Ned" Jackson Professor and past Director, MSU Center for Fundamental Materials Research. A.B., 1977, Harvard University; Ph.D., 1987, Princeton University; Postdoctoral Fellow, 1986-88, Ohio State University. Physical Organic Chemistry. Reactive intermediates and reaction mechanisms; hydridic-to-protonic hydrogen bonding (AKA "dihydrogen bonding" organic magnetic and conductive materials; catalytic pathways to chemical intermediates.  Tel. (517)-355-9715 ext. 141.  E-mail: jackson@cem.msu.edu Vita and Publications

Hydrogen-Hydrogen Hydrogen Bonding

    In work started in summer of 1994, and primarily carried forward by summer high school students and undergraduates, we demonstrated an interaction between the electron pair of a B-H bond and a traditional H-X (X=N, O, Halogen) hydrogen bonding partner. Single crystal X-ray and neutron diffraction, solution spectroscopies (IR, NMR), and ab initio calculations all indicate that such relationships are best understood as a novel type of hydrogen bonding.

    While our own studies have focused on the borohydrides, related examples of close interactions between hydridic and protonic hydrogen sites in organoiridium complexes have been reported from the Crabtree group at Yale. [Crabtree, R. H.; Siegbahn, P. E. M.; Eisenstein, O.; Rheingold, A. L.; Koetzle, T. F. Acc. Chem. Res. 1996, 29, 348-354.]

   The new associative effect has implications for regiochemical control of organic reactions and for the rational assembly of crystalline, covalent materials. Consistent with calculations (shown at left) Sterling Gatling has uncovered strong, controllable directing and rate effects in borohydride reduction of hydroxyketones under appropriate conditions.[50] The idea of chemical "basting stitches"-weak linkages that can organize and hold a structure's form while it is more firmly sewn together-is also very appealing. Hydrogen-hydrogen H-bonding represents just such an interaction, and Radu Custelcean's work in this area amply demonstrated the idea. He established that loss of H₂ from HHH-bonds in the solid state can lead to products distinct from those in fluid phases (melt or solution) and that in favorable cases, crystallinity can be maintained throughout the reaction despite some geometry change and the release of gas inside the crystals [45,47]. It is now clear that HHH-bonding can be used as a key element in crystal engineering. Because of their reliable structural preferences and compactness, HHH-bonds may serve as key elements in strategies aimed at synthesis of new covalent linkages in solid crystal lattices [52,54]. We have recently reviewed the chemistry of this interesting interaction [57], and presented our first efforts to assemble extended covalently bonded networks via the above HHH-bond-directed pathways.[60, 62].

     The hydridic-to-protonic hydrogen bonding project, begun in the 1994 with high school student Mani Sharma, has benefited from the participation of several talented and hardworking high school students--Mani Sharma ('94), Charley Wilson ('95), Susanna Bass ('96), and Jennifer Hung ('97). Three of the four entered the Westinghouse Science Talent Search competition, and two (Sharma and Bass) were semifinalists, among their many awards. The X-ray structure of guanidinium borohydride, an interesting example of multipoint HHH-bonding, came out of Susanna Bass's studies in the summer of 1996. After Drs. Gatling and Custelcean moved on to new horizons, Simona Marincean, aided by such talented undergraduate researchers as Gates Cambridge Scholar Robin Stein, Melinda Baker, and Justin Roberts (all now attending top ten graduate schools in Chemistry), uncovered new segments of this unusual story. Simona, now Dr. Marincean, used ab initio methods to study the dynamics of proton transfer in the reaction of various acids with BH₄ and AlH₄ anions. [71] Together with Custelcean and Stein, she also reexamined the crystal structure of NH₄H₂PO₂, a salt whose crystal structure, reported in 1934 before the "classical" hydrogen bond had even been recognized as a common structural theme. This compound's crystal structure had been analyzed in terms which implied hydridic-to-protonic hydrogen bonding, but though their insight turned out to be propetic, our modern structural reanalysis found no such bonding in this compound. [74]
 

Oxygen Atom Transfer Reactions of Carbenes:

    Carbenes' most familiar reactions are insertions into C=C pi and R-H sigma bonds, and H or halogen abstractions. However, we have recently begun to systematically examine O atom abstraction (see below) by the reactive carbenes fluorenylidene (Fl:) and methylene (H₂C:, the simplest carbene).

    That C atoms could abstract oxygen from nearly anywhere has been known for some time. Although the reaction is far less exothermic, we now know that fluorenylidene and methylene (Fl: and H₂C:) are similarly able to deoxygenate suitable carbonyl compounds to yield a new carbene and the carbonyl products fluorenone or formaldehyde (Fl=O or H₂C=O).[35] This atom transfer, which breaks and forms C=O double bonds, represents a rare reaction path for carbenes. The overall process may vary from near thermoneutral to highly exothermic, and may have an observable intermediate or none. In strongly exothermic cases, as with the tetramethyl urea (TMU) substrate shown above, it can be very rapid (4 x 10⁸ l/mol.sec). We are pursuing experimental and theoretical studies to establish the scope, kinetics, and thermodynamics of these reactions and to ask whether they are one- or two-step processes. Recent results of Dalila Kovacs' ab initio calculations (MP2/6-31G* level) on this system point to the formation of an ylide intermediate as shown at right, despite the large exothermicity of the overall reaction. Thus, it appears that even this rather favorable case proceeds stepwise.

    Dalila's hig level ab initio theoretical studies of the closely related reaction H₂C: + CO₂ --> H₂C=O + CO have uncovered a single-step oxygen abstraction pathway with an activation barrier of 24 kcal/mol on the singlet potential energy surface.[56] However, in the most favorable path, methylene attacks CO₂ without an enthalpic barrier to form oxiranone (aka alpha-lactone), a 47 kcal/mol exothermic reaction. This product then traverses a second barrier, 20 kcal/mol below the energy of the starting species, to fragment into CO and CH₂O, for an overall release of 61 kcal/mol. We believe that the initial addition reaction produces "hot" alpha lactone with its vibrational energy in just the right modes to rapidly continue over the fragmentation barrier to form CH2O + CO. The dynamics of such a process would be interesting and amenable to computational modeling, which is currently getting underway. Here is a "movie" from the dynamic reaction coordinate (i.e. total energy fixed oscillating back and forth between kinetic and potential forms) for CH₂+CO₂ --> CH₂O + CO as calculated at the relatively primitive HF/STO-3G level. Some of the frames from the movie may be viewed as an animated or static "slide show."
 

Symmetry Analysis of the Carbene-Olefin Cycloaddition Transition Structure
 
 

As noted above, one of the most familiar reactions of carbenes is insertion into C=C double bonds to form cyclopropanes. Based on the theoretical studies by Hoffmann and Dewar in the early '70s, this process has long been "known" to occur via an unsymmetrical reaction path wherein the carbene tilts to orient its empty 2p orbital to attack the electron pair of the pi bond, while the carbene's sp2 lone pair remains "out of the way" until the reaction has progressed well along the reaction coordinate.  This view of the reaction, founded on orbital symmetry arguments, remained experimentally unexplored for roughly two decades until Julius Remenar's studies here at MSU.  Julius applied "Tolbert analysis" [ Tolbert, L. M.; and Ali, B., J. Am. Chem. Soc. 1981, 103, 2104] to the reaction of :CCl2 with three dialkyl (diethyl-; l-bornyl, ethyl-; and di-l-bornyl-) fumarates, uncovering evidence that supports the asymmetric trajectory.

    The activation energy difference between attack on the two alkene faces is of course zero for diethyl fumarate (faces are enantiotopic); replacement of one ethyl group by bornyl introduces a nonzero difference, which would be doubled for the dibornyl case, if the path is symmetrical. Thus, the Tolbert method predicts that for a symmetrical approach to the olefin, the diastereofacial preference for attack on dibornyl fumarate should be the square of that on the bornyl, ethyl substrate. The observed values are 1.21+/-0.3, and 1.28+/-0.04 for the bornyl,ethyl and dibornyl cases, respectively; the latter value falls solidly and reproducibly below the 1.4-1.5 range predicted for a symmetric path. Interestingly, semiempirical (PM3) molecular orbital calculations for the specific system studied do find an asymmetric reaction path, but their predicted diastereoselectivity ratios (the basis for the Tolbert method's symmetry assessment) are much smaller and closer to the values expected for a symmetrical path than we actually found experimentally.
 

Self-assembly of Magnetic Materials:

(RM)_n oligomer

    Magnetism arises from the cumulative behavior of unpaired electrons, so design of a molecular magnetic material requires understanding and ultimately control of the system's spin-couplings. One ion-binding strategy for self-assembly that defines structural and magnetic relationships between organic radicals is shown below, followed by pieces--molecular models we have built--of the (RM)_n magnetically coupled chain. In the left X-ray structure, two radicals flank a bound K⁺ ion ("RMR"); in the structure at right, from pulsed EPR spectroscopy and computer modeling, two Li⁺ cations flank a radical ("MRM").[24] From detailed structural and spectroscopic studies of diamagnetic analogues [28, 38], and by examining related systems' magnetic behaviors [30, 31] we are building an understanding of the rules relating structure to electron coupling. We have shown that new couplings can be introduced by ion binding; continuing work in new structural motifs has begun to target extended chains and control of interchain coupling in "simple" salts [36] and hydrogen bonded systems [48] as well as magneto-structural correlations in layered vanadyl phosphate and organophosphonate magnetic materials.[32, 41] We are also interested in the magnetic characteristics of alkali metal salts of organic radical anions, of which we have synthesized several sets.[46,53] After careful consideration, we have designed a new diradical dianion building block for organic magnetic materials related to ortho-semiquinone ligands.[55] This work is largely carried out in the context of a rich collaboration with Prof. James L. Dye, the discoverer of alkalides and electrides, salts in which the anions are alkali metal anions or simply electrons, charge-balanced by alkali metal cations bound in neutral complexants such as crown ethers or cryptands. Such compounds are notoriously sensitive to heat and air, and we have recently been working to develop ligands that were more highly resistant to reduction than the crown ether/cryptand classes. Our efforts have paid off in the synthesis of the first room temperature stable alkalides [51], the first alkalide in which the cation is a complexed proton (H⁺) [63], and a novel six-coordinate Li complex in a peraza cryptand.[66]
 

Mechanisms of Catalytic Processes:

• Biomass-derived Organic Acids --> Useful Chemicals

• Methanol --> Gasoline Hydrocarbons

Other interests include:

Computational Modeling of Carbene and Radical Reaction Paths;
Solid State Reaction Mechanisms;
Strained Hydrocarbons