Eight is not enough | National Center for Supercomputing Applications at the University of Illinois

Eight is not enough

09.09.11 -

by Trish Barker

A team of Illinois chemists discovers that a previously unknown type of bond is responsible for the stability of molecules with more bonds than the Octet Rule predicts.

"How atoms bond is about as fundamental as it gets for chemistry," says Illinois chemistry graduate student Jeff Leiding.

Given that truism, you might think that chemists would have figured out everything there is to know about the phenomenon since Isaac Newton outlined a theory of atomic bonding in 1704. But until recently, chemists weren't sure how to describe the bonding in the large class of molecules that form more than the expected number of bonds—so-called hypervalent molecules. Using the Abe and Ember supercomputers at NCSA in addition to computational resources in the chemistry department, the Illinois research team led by NCSA director Thom Dunning and research chemist David Woon has discovered a new type of bond that explains hypervalency and has published several explorations of this new mode of bonding since 2009.

"Chemists have long known that the chemistry of the main group elements nitrogen through fluorine is different than that of the elements in succeeding rows of the periodic table. The question was: What makes phosphorus different than nitrogen or sulfur different than oxygen? It appears that this new type of bond is a major cause of this anomaly," Dunning says. "This phenomena has implications for all of chemistry."

A common anomaly

For decades students have been taught the Octet Rule, which states that atoms want to have eight electrons in their valence shells. When you know how many electrons an atom has in its outer shell, you know how it will gain, lose, or share electrons to form complete octets, which tells you the number of bonds the atom can form.

The real world of molecular bonding doesn't neatly conform to this rule, however. Hypervalent molecules form more than the expected number of bonds, appearing to stuff their valence shells with more than eight electrons. The greenhouse gas sulfur hexafluoride (SF6) is one example; another is chlorine trifluoride (ClF3), a gas used in rocket fuel, nuclear fuel processing, and etching operations in the semiconductor industry. It turns out, many molecules break the Octet "Rule."

"It's really common behavior," Leiding says. "There's been this entire class of compounds that has been poorly understood in terms of how they bond and how they react."

Through the years, scientists had proposed several theories to explain hypervalency, but none provided a satisfactory explanation. Then in 2009, Woon and Dunning conducted computational studies (using the Molpro suite of quantum chemistry programs) of the ground and low-lying excited states of the sulfur fluorine species: SF, SF2, SF3, SF4, SF5 and SF6.

"Building molecules one atom at a time provides unrivaled insights into the nature of bonding," they wrote in The Journal of Physical Chemistry A. The article outlined their theory that hypervalency is caused by a previously unknown type of chemical bond, which they dubbed a "recoupled pair bond."

Imagine two atoms, one with two electrons in an outer orbital and the other with just one. Normally, the paired electrons wouldn't participate in a bond. In order for a bond to form, the pair must be split apart. Some atoms, like fluorine, are able to force that split. One electron from the original pair is "recoupled" by the fluorine, forming a recoupled pair bond with the electron in the singly occupied fluorine orbital. The other previously paired electron is now available to form another bond. Two bonds can form where you would expect none.

"The model explains so much that had previously been unexplained in these compounds," Leiding says, such as the "wildly oscillating bond energies of the sulfur fluoride series" and the presence of unexpected low-lying excited states. "The overarching thing about [recoupled pair bonding] is that it describes so much behavior that was thought to be anomalous. And it's really not anomalous at all. This behavior is really rampant in the third row and beyond in the periodic table," he says.

The team has even found recoupled pair bonding in "normal valent" molecules such as SF2, albeit in excited rather than ground states. In addition, much of the chemistry of the early elements in each row (even in carbon) can be explained by recoupled pair bonding.

Predicting elusive isomers

The team has continued to investigate recoupled pair bonding in phosphorous and chlorine halides as well as the oxides and hydroxides of sulfur (which are significant in understanding atmospheric chemistry, such as acid rain) and the noble gas halides.

"The nice thing about our theory," Woon says, "is that it is a predictive theory. When we studied compounds of phosphorous and chlorine, we had much better intuition about how they would behave based on what we'd learned about recoupled pair bonding in sulfur compounds."

The team's most recent work, published this spring in The Journal of Physical Chemistry A, uses recoupled pair bonding to predict and calculations to verify the existence of long elusive "bond stretch" isomers, compounds that have the same molecular formula but different structures due just to variations in bond lengths. The Illinois team's computational study of the SFn-1Cl series found that, in a low-lying excited state, SFCl can exist with a covalent bond between S and F and a recoupled pair bond between S and Cl or with the bonds reversed (a recoupled pair bond between S and F and a covalent bond between S and Cl).

"The recoupled pair bonding model successfully predicted the existence of several isomers that might otherwise be overlooked," they wrote. While the SFCl isomers won't exist as independent species, the vibrational structure of the molecule, which has not yet been observed in the laboratory, will provide clear evidence for their existence.

In addition to further exploration of recoupled pair bonding, the team plans to rewrite the book on bonding.

"We looked at the general chemistry textbooks to see how they are teaching bonding in general and said...'Aaargh!'" Woon says. Computational chemistry and modern experiments have provided profound new insights into the nature of the chemical bond. "We have funding from NSF to update the way that the subject of bonding is covered in introductory chemistry courses."

For more information, go to: http://chemistry.illinois.edu/faculty/Thom_Dunning.html.

Team members
Lina Chen
Thom Dunning
Jeff Leiding
Beth Lindquist
Tyler Takeshita
David Woon
Lu Xu