Most KS3 chemistry tutoring I've seen goes like this. Open the textbook. Find a worked example. Walk the child through balancing the equation. Repeat with three more. Move on.

That works fine right up until Year 10, when the curriculum starts assuming things the child never actually understood. Mole calculations. Empirical formulae. Why concentrated acid and strong acid aren't the same thing. The maths the child has been practising for three years suddenly stops behaving, and no one can quite say why.

Prof Curie is built to prevent that. She teaches the conceptual picture before the numerical one, every time, deliberately. It's slower for the first fortnight. It's enormously faster for the next three years.

The first principle

If you ask Curie what her job is, the answer is short. Before any chemistry happens at all, she's checking whether the student has the right mental model of what particles are doing. She does this even on topics that look procedural — balancing equations, calculating pH, identifying reaction types.

The reason is that KS3 chemistry is, at its core, a story about invisible things. You can't see ions. You can't see bond breaking. You can't see oxidation. The whole subject is a translation exercise between three layers: what the student observes, what the particles are doing, and how we write it down with symbols.

If the particle layer is wrong, the symbol layer is just arbitrary scribbles. The student can pass tests for a year or two by memorising the scribbles. Then it stops working.

What a Curie session actually looks like

The opening of a topic is never a definition. It's a question that probes for the wrong model. For example:

  • "When sugar dissolves in tea, where do the sugar particles go?"
  • "In a sealed jar, you let iron rust. Does the mass change?"
  • "When we heat a gas, do the particles get bigger?"

These are diagnostic. A student who says "the sugar disappears" or "the gas particles expand" is showing Curie exactly where to start. A student who says the right thing gets the same question back the other way ("good — and how would you convince a friend who didn't believe you?").

The whole topic then unfolds from that opening. Macro: what do you actually see? Submicro: what are the particles doing? Symbolic: how do we write it? Only at the symbolic layer do numbers start to appear.

Three reasons this matters more in chemistry than other sciences

Physics has the same three-layer thing, but the observations and particles are usually the same level — you can see a ball, and the ball is the object you're computing on. Biology is more visual than chemistry; cells and organs are at least drawable. Chemistry is the subject where the observation and the particles are radically different scales. You watch a colour change; the explanation is at the scale of nanometres. That mismatch is where misconceptions form.

The second reason is that chemistry vocabulary front-loads early and is unusually unforgiving. Ion, isotope, displacement, oxidation, mole — five words that, if any is wobbly, makes a sentence in a GCSE exam unparseable. (Curie has a whole article on the vocabulary problem.)

The third reason is the maths. Stoichiometry is, on the surface, ratio arithmetic — material a competent KS3 maths student can do. But the meaning of those ratios is particle behaviour. A student who treats stoichiometry as ratio puzzles without understanding what 2 moles of magnesium means is doing one of those exercises where the answer is right and the understanding is hollow.

How Curie differs from textbook order

A KS3 chemistry textbook will typically introduce balanced equations in Year 7 or 8, with the maths first and the particle interpretation second. Curie deliberately reverses this. By the time a student in her sessions is balancing equations, they've already drawn the reaction at the particle level, counted atoms in a picture, and only then met the coefficients-in-front-of-formulae version.

It's the same destination. The order is different. Order matters.

Tone modes

Like every faculty agent, Curie adapts to the weekly energy level the Mentor sets. On a green week she'll lean experimental — "what if we tried...", harder probes, GCSE preview. On amber she'll slow down, more real-world anchoring, fewer new concepts. On red she's homework support only, no new learning at all.

The chemistry-specific signal she watches for is the speed of recall. A student who can name an ion confidently and quickly is at one level. A student who knows the answer but takes ten seconds to say it is showing Curie that the model is present but fragile. That's a sign to slow down, not push forward.

What she won't do

A few firm rules.

  • She won't give a chemistry answer before checking the underlying model. Even on a "quick question" she'll usually ask one diagnostic before answering.
  • She won't run real practicals. She'll talk through them, reference safe kitchen versions, and connect what the student saw at school to the three-layer picture. The school's lab is the school's lab.
  • She won't replace a chemistry teacher. KS3 chemistry has structural content that needs sequencing across two or three years. Curie supports that sequencing; she doesn't drive it.

What to expect as a parent

If your child has had a Curie session and you ask "what did you do?", you probably won't hear about a worked example. You'll hear something like "she kept asking me what the particles were doing". That's the signal that the session was working as designed.

The reverse signal — "she gave me five practice questions" — would mean the agent had slipped into homework-helper mode. Curie's design pushes hard against that. If your child says the sessions feel "slow at first", that's also the design working. The catch-up happens around session six or seven, when topics that would normally take a week to land click in twenty minutes because the model underneath is solid.

A note on the named professor

I should say something about the name. Marie Curie is one of the more remarkable scientists in history — twice a Nobel laureate, in two different sciences, in an era that did not welcome women in either field. Naming a tutoring agent after her isn't decoration. The choice is partly because Year 7-9 chemistry is the topic students most often decide isn't for them, often before they've actually met it properly. A small reminder, in the name on the screen, that the discipline isn't an old boys' club seems like a quiet good thing.

She also worked entirely with experimental, model-driven chemistry, with vocabulary that didn't yet exist when she started. That happens to be exactly what the agent does too.

The other reason for the name is more pedagogical. Curie's actual scientific style — patient, hands-on, willing to look stupid in front of an unexpected result — is the right disposition for KS3 chemistry too. A 12-year-old who's scared of being wrong won't try a probe question. A 12-year-old who's seen that being wrong is part of the process will. That second disposition is what the agent is, quietly, trying to build.

What's coming in the chemistry cluster

This is the overview piece. Two follow-on articles go deeper into the method:

  • The next piece walks through how Curie uses Bohr atoms, particle diagrams, and equation skeletons to make a topic land before any numbers appear.
  • The piece after that covers the KS3 chemistry vocabulary problem — the words that, if they're shaky, make GCSE chemistry feel like a foreign language.

If your child is starting KS3 chemistry or is already feeling lost in it, those two articles together give a clearer picture of what the agent will actually do session by session.

For parents new to the way KS3 chemistry is structured in England, BBC Bitesize is a useful free reference for what your child should be meeting in each year, and the Royal Society of Chemistry's teacher resources are unusually good for non-specialist parents who want to follow along.

Jason runs aitutors.me. He has a Year 8 child and about fifteen years of building software adjacent to education. Updated 21 May 2026.