What a curly arrow really claims
What a curly arrow really claims
A-level Chemistry draws reaction mechanisms as though someone had watched them happen. Nobody has. A mechanism is an inference assembled from indirect evidence, and knowing how chemists build one separates doing the subject from memorising it.
Organic chemistry at A-level is taught through the curly arrow. An arrow leaves a lone pair or a bond, lands somewhere else, and a reaction is explained. Students learn to draw these fluently, and we have argued before that understanding the reasoning behind them beats memorising them, in our note on the Chemistry topics students quietly underestimate. What is rarely pointed out is stranger and more interesting. Nobody has ever seen a curly arrow happen. The mechanisms in your textbook are not observations. They are arguments.
What a mechanism actually is
A mechanism proposes a sequence: which bonds break, which form, in what order, and what fleeting structures exist in between. Those structures are the difficulty. A transition state is not a substance you could isolate. It is the highest point on the path from reactant to product, and it exists for a span measured in femtoseconds. The carbocation at the heart of an SN1 reaction survives for nanoseconds at best. When your textbook draws one, it is drawing something no chemist has ever held in a flask. A mechanism is a reconstruction of an event that finished long before any instrument could look at it.
How chemists build the case
The evidence is entirely indirect, and gathering it is where the discipline turns clever. Kinetics came first. If the rate of a substitution depends on the concentration of the nucleophile as well as the halogenoalkane, both must be present in the slow step, which is the SN2 picture. If the rate depends on the halogenoalkane alone, the slow step involves that molecule by itself, which points to a carbocation forming before the nucleophile arrives. Stereochemistry supplies a second line of attack: when a reaction at a chiral carbon reliably inverts the configuration, the nucleophile must be approaching from the side opposite the leaving group, which is what one mechanism predicts and the other does not. Isotopic labelling supplies a third. Swap a hydrogen for a deuterium, and if the reaction slows measurably, that bond is breaking in the rate-determining step. No single result reveals a mechanism. Taken together they narrow the field until one proposal survives and its rivals do not.
Proposals get revised
Because a mechanism is a hypothesis, it can be overturned, and plenty have been. The clean division between SN1 and SN2 that A-level presents as two separate boxes turned out to describe the two ends of a continuum, with a wide borderline region where reactions proceed through ion pairs that are neither fully separated nor properly bonded. In the 1990s Ahmed Zewail came closer to the transition state than anyone had managed, using laser pulses brief enough to track molecules partway through a reaction, work that won the 1999 Nobel Prize in Chemistry and opened the field of femtochemistry. Even that does not photograph an arrow. It follows how a population of molecules changes over the time a bond takes to break, which is as near to direct observation as the subject has come.
The notation is not the molecule
A related warning applies to the other things you draw. Benzene is the clearest case. Students are shown two Kekule structures with a double-headed arrow between them, and many come away believing the molecule flickers between the two. It does not. Benzene is one unchanging structure with six identical bonds, and the pair of drawings exists only because a notation that insists every bond is either single or double cannot capture it in a single picture. Resonance is a limitation of our language, not a behaviour of the substance. Roald Hoffmann, who has written more carefully than anyone about what chemical diagrams do and do not mean, treats this as central to the subject rather than a footnote to it.
Why this changes how you work
The practical payoff arrives quickly. Treat mechanisms as arguments rather than facts to recall and unfamiliar questions become tractable. If you know that arrows run from electron-rich to electron-poor, that charge has to balance at every step, and that any intermediate you propose must be something capable of existing, you can construct a mechanism for a molecule you have never met and defend it. That is a different activity from remembering one, and it is what the harder questions are built to test. It also explains why chemistry interviewers so reliably produce a reaction that appears on no syllabus. They are not checking your recall. They want to watch you reason about something invisible from the evidence in front of you.
Seen this way, chemistry sits alongside the rest of the sciences in a manner the syllabus rarely makes clear, and it is the same argument we make about physics and its deliberately false models. Peter Atkins on what happens inside a reaction, and Hoffmann's essays on chemical thinking, are good places to begin, and both sit on our reading lists. Building this habit of reasoning is what our Chemistry tutoring and admissions support is for.
