Discovery of ‘knock-on chemistry’ opens new frontier in reaction dynamics

New experimental findings cast in doubt 90-year-old theoretical model of the transition state in chemical reactions


Research News


TORONTO, ON – Research by a team of chemists at the University of Toronto, led by Nobel Prize-winning researcher John Polanyi, is shedding new light on the behaviour of molecules as they collide and exchange atoms during chemical reaction. The discovery casts doubt on a 90-year old theoretical model of the behavior of the “transition state”, intermediate between reagents and products in chemical reactions, opening a new area of research.

The researchers studied collisions obtained by launching a fluorine atom at the centre of a fluoromethyl molecule – made up of one carbon atom and three fluorine atoms – and observed the resulting reaction using Scanning Tunneling Microscopy. What they saw following each collision was the ejection of a new fluorine atom moving collinearly along the continuation of the direction-of-approach of the incoming fluorine atom.

“Chemists toss molecules at other molecules all the time to see what happens or in hopes of making something new,” says Polanyi, University Professor in the Department of Chemistry in the Faculty of Arts & Science at U of T and senior author of a study published this month in Communications Chemistry. “We found that aiming a reagent molecule at the centre of a target molecule, restricts the motion of the emerging product to a single line, as if the product had been directly ‘knocked-on’. The surprising observation that the reaction product emerges in a straight line, moving in the same direction as the incoming reagent atom, suggests that the motions that lead to reaction resemble simple onward transfer of momentum.

“The conservation of linear momentum we observe here suggests a short-lived “transition state”, rather than the previous view that there is sufficient time for randomization of motion. Newton would, I think, have been pleased that nature permits a simple knock-on event to describe something as complex as a chemical reaction,” says Polanyi.

The team, which included senior research associate Lydie Leung, graduate student Matthew Timm and PhD graduate Kelvin Anggara, had previously established the means to control whether a molecule launched towards another either collides head-on with its target or misses by a chosen amount – a quantity known as the impact parameter. The higher the impact parameter, the greater the distance by which the incoming molecule misses the target molecule. For the new work, the researchers employed an impact parameter of zero to give head-on collision.

“We call this new type of one-dimensional chemical reaction ‘knock-on’, since we find that the product is knocked-on along the continuation of the direction of reagent approach,” says Polanyi. “The motions resemble the knock-on of the steel balls of a Newton’s cradle. The steel balls of the cradle don’t pass through one another, but efficiently transfer momentum along a single line.

“Similarly, our knock-on reactions transfer energy along rows of molecules, thereby favouring a chain-reaction. This conservation of reaction energy in knock-on chemistry could be useful as the world moves toward energy conservation. For now, it serves as an example of chemical reaction at its simplest.”

It has been known for well over a century that there is an energy barrier that chemical reagents must cross on their way to forming reaction products. An energized transition state exists briefly at the top of the barrier in a critical configuration – no transition state, no reaction.

Polanyi says the observation of collinear ‘knock-on’ provides insight into the reactive collision-complex, which lasts for approximately a million-millionth of a second. “Our results clearly tell us that the transition state at the top of the energy barrier lasts for so little time that it cannot fully scramble its momenta. Instead, it remembers the direction from which the attacking fluorine atom came.”

In the 1930s, chemists began calculating the likelihood of forming a transition state on the assumption that it scrambles its energy, like a hot molecule. Although it was an assumption, it appeared well-established and gave rise to the statistical “transition state theory” of reaction rates. This is still the favored method for calculating reaction rates.

“Now, with the ability to observe the reagents and the products at the molecular level, one can see precisely how the reagents approach and subsequently how the products separate,” Polanyi says. “But this runs contrary to the classic 90-year old statistical model. If the energy and momentum were randomized in the hot transition state, the products would not exhibit a clear memory of the direction of approach of the reagents. Energy-randomization would work to erase that memory.”

The researchers say the observed directional motion of the reaction products favours a deterministic model of the transition state to replace the long-standing statistical model. Additionally, the observed reaction dynamics allow the reagent energy to be passed on in successive collinear collisions.


The research was funded in part by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the University of Toronto NSERC General Research Fund. Theoretical calculations were performed on the Niagara cluster at SciNet HPC Consortium.
From EurekAlert!

5 8 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
March 15, 2021 10:56 pm

Not so fast there eurekaAlert. The original research states that 25 out of 71 experiments showed no knock-on lineal direction continuity with striker’s direction, with 46 out of 71 experiments showing the knock-on directionality that the original post highlights.

As the published research clearly states, in fact, 35% of the non performing knock-ons’ displacement was in the opposite direction from incoming striker. Their supplemental material discusses this in more detail.

Reply to  gringojay
March 16, 2021 1:24 am

35% of skeptics resisted. Just move them to the edge of the table, slide them off. This the new science.

Reply to  gringojay
March 16, 2021 6:23 am

Pool balls which collide precisely head-on with respect to each balls center of mass do not scatter energy; they transfer it along the path of travel. Those which collide at any angle other than precisely head-on do scatter energy. Minnesota Fats knew all about this. What these researchers seem to be saying is that collisions between fluorine atoms shot from some device towards trifluoromethane molecules occur precisely head-on and conserve linear momentum 65% of the time. So either their aiming mechanism for the “fluorine gun” is extraordinarily accurate or there is some phenomena that makes it appear that way – 65% of the time. I wonder which it is.

Bryan A
Reply to  DHR
March 16, 2021 6:38 am

Newton’s Cradle on the untethered atomic level

Last edited 26 days ago by Bryan A
Clyde Spencer
Reply to  DHR
March 16, 2021 8:07 am

Coincidentally close to the 1-sigma probability distribution?

Reply to  DHR
March 16, 2021 4:45 pm

Or the fluorine atom rebounded 35% of the time.

Sounds exactly like conservation of energy as demonstrated in pool rooms everywhere.

Straight pool or nine ball?

Reply to  DHR
March 16, 2021 5:04 pm

At the atomic level, perhaps uncertainty. The more precise the aim/position, the less precise the knowledge of the momentum.

Reply to  gringojay
March 16, 2021 8:28 am

And it’s crock of rubbish by some chemist trying to understand something out of his grasp.
The nobel prize was awarded in 1999 for imaging chemical reaction states using using femtosecond spectroscopy (

SLAC has done thousands of reactions and even has movies of them since 2015

It is exceedingly rare for a chemical interaction to maintain classical properties.

About all we got from the work was how out of touch with modern research one chemist is.

Reply to  LdB
March 16, 2021 11:41 am

Here’s the original sound track:

“Well I’m rollin’ & I’m tumblin’

When I woke up this mornin’,
I couldn’t tell right from wrong

Well if the river was whisky,
I’d be a diving duck

I’d dive down to the bottom,
I’d never come up ….”

March 15, 2021 11:13 pm

What is this heresy !
“Observations run contrary to the classic 90-year old statistical model”
“experimental findings cast doubt in theoretical model”
wash your mouth out, you can’t go around saying things like that !

Models are the epitome of perfection which is why we run our flawless society with them …oh wait a minuet …

Besides, a 90-year old model probably has Alzheimer’s & will be unreliable, so shouldn’t be questioned.

Last edited 26 days ago by saveenergy
Bryan A
Reply to  saveenergy
March 16, 2021 6:39 am

Our society runs on the Whims of Supermodels

March 15, 2021 11:38 pm

In Chemical engineering there are similar equations with dimensionless variables for heat transfer, mass transfer and momentum transfer. This is not new. Science lags engineering knowledge but science is good at thought bubbles within their limited knowledge just as with the Climate scam where no so-called scientist takes note of engineering thermodynamics.

Reply to  Proeng
March 16, 2021 12:15 am

Kezactly! As one reads this (admittedly short and redacted) report, the question that comes to mind is: “Yeesss, but is it art?”

March 16, 2021 12:12 am

Obviously they’re not climate scientists. Fancy rejecting a 90 year old model for some observations!

March 16, 2021 4:47 am

I lost interest when a trifluoromethyl group was called a molecule.

Loren C. Wilson
Reply to  Bsl
March 16, 2021 5:17 am

The drawing shows a carbon atom (black), a hydrogen atom (white), and three fluorine atoms (green), which makes trifluoromethane, which is a molecule. All bond requirements are satisfied, no free radicals or open valence pairs. You used to be able to buy this in bulk.

Reply to  Loren C. Wilson
March 16, 2021 6:38 am

I saw the diagram. My objection is the text:
the centre of a fluoromethyl molecule” 

Steve Z
Reply to  Loren C. Wilson
March 16, 2021 9:17 am

In the diagram, it appears that the black circle represents carbon, green circles represent fluorine, and white circles represent hydrogen. So it appears that a fluorine atom is “launched” at a molecule of trifluoromethane (CHF3), there is a brief “excited” state of CHF4-, which then emits another fluorine atom, while the trifluoromethane molecule is unchanged. This is not a chemical reaction, so it is not surprising that momentum is conserved.

The diagram also shows other white circles (hydrogen atoms?) sitting there, which would likely react with fluorine to form hydrogen fluoride (HF), which can also polymerize to form short chains (HF)n.

Fluorine is the most electronegative element in the entire periodic table, so that if a reaction occurred, fluorine would replace the hydrogen atom to form carbon tetrafluoride (CF4) and emit a hydrogen atom. It is not clear that linear momentum would be conserved, due to the energy of the reaction, and there could be rotational effects, so the hydrogen atom may be emitted in a different direction.

What is also surprising is the “launching” of free fluorine atoms at trifluoromethane molecules. Elemental fluorine does not exist in nature as single atoms, but as two fluorine atoms covalently bonded (diatomic gas F2). It takes a lot of energy to split a fluorine F2 molecule into separate atoms, which are extremely reactive, and would be difficult to “aim” at a trifluoromethane molecule without being ionized. If free hydrogen atoms were present (or even H2 molecules), a free fluorine atom would be more likely to react with hydrogen (to form HF) than with trifluoromethane.

If a free fluorine atom was “launched” at a trifluoromethane molecule, the result would depend on the orientation of the CHF3 molecule, which in the gas phase is in constant rotation. Trifluoromethane has a tetrahedral (triangular pyramid) structure, with the fluorine atoms and hydrogen atom at the vertices, and the carbon atom at the center. It is also a polar molecule, with the fluorine atoms negatively charged, and the hydrogen atom positively charged.

If the “launched” free fluorine atom approached one of the fluorine atoms of the CHF3 molecule, it would be repelled by the negative charge, and effectively “bounce” away, and momentum would be conserved. If it was launched at the hydrogen atom, it would be attracted by the positive charge, and likely react to form carbon tetrafluoride (CF4), and momentum may not be conserved, due to the energy of reaction.

Don Bennett
March 16, 2021 5:52 am

Well, it’s also been know for some time that the “coulomb barrier” does not need to be breached to have a reaction take place. I remember well the class discussion of “quantum tunneling” whereby reactions take place well below the level of conditions thought needed for the reactions to procede.

March 16, 2021 6:42 am

Unfortunately, this stuff is “old hat” now.
Advances in understanding the forces holding molecules together has transitioned (for some chemists) into an understand of the picometric vibrations that result from the Casimir force interactions. Note that the following formative article is dated 2009:

I am of the “bang-molecules together” generation, but many chemistry departments are studying catalysis from the Casimir aspect.

Jaap Titulaer
March 17, 2021 2:02 am

In the 1930s, chemists began calculating the likelihood of forming a transition state on the assumption that it scrambles its energy, like a hot molecule. Although it was an assumption, it appeared well-established and gave rise to the statistical “transition state theory” of reaction rates. This is still the favored method for calculating reaction rates.

The researchers say the observed directional motion of the reaction products favours a deterministic model of the transition state to replace the long-standing statistical model.

Well not exactly. Statistics still play a big role, just at a different level (above the single reaction, not inside, or at the level of, the single reaction).

While each single reaction may be the result of an atom moving in a straight line (and deterministic), the overall reaction still involves multiple molecules hitting multiple target molecules from all directions (and chaotic).

How chaotic that is depends (a.o.) on the temperature. It remains something where statistics can be applied, simply because there are numerous molecules involved, all moving in different directions at different speeds.

It’s only at the single individual molecular reaction level that the chaotic assumptions can be dropped.

Tom Schaefer
March 18, 2021 8:41 am

Is microwave activated chemistry a thing? I would think it would be a pretty easy thing is to identify the energetic vibrational states, and the microwave frequency to excite them, that would facilitate such reactions. Back when we were thinking about putting large HF or DF chemical lasers in space for BMD, that was the process.

Last edited 24 days ago by Tom Schaefer
%d bloggers like this: