The Arrhenius Equation: Why Temperature Rules Everything

Three articles in, and one symbol has been quietly doing all the work while dodging every question: $k$.

Part 3 called it “the rate constant” and moved on. But $k$ is only constant in one specific sense — it doesn’t depend on concentration. It depends ferociously on temperature. And “ferociously” is the right word, because the dependence is exponential, which means your intuition about it is almost certainly wrong.

The equation

Three pieces, and each one means something physical:

• $E_a$ — activation energy (J/mol). The energy barrier a collision must clear for reaction to happen. Molecules don’t react just by bumping into each other; they have to bump hard enough.
• $A$ — the pre-exponential factor. Roughly, how often collisions happen at all, with the right geometry. It sets the ceiling: $k$ can never exceed $A$.
• $e^{-E_a/RT}$ — the fraction of collisions with enough energy. This is the term that does all the damage. It’s the fraction of molecules out in the high-energy tail of the distribution.

Read the equation as a sentence: rate constant = (how often molecules collide) × (the fraction of those collisions that are energetic enough to matter). Raising temperature doesn’t make molecules collide much more often. It makes far more of those collisions violent enough to count.

That distinction matters. Heating a gas from 300 K to 310 K increases the average molecular speed by well under 2%. Collisions barely get more frequent. But the tail of the energy distribution — the molecules with enough punch to clear the barrier — grows dramatically. All the temperature sensitivity lives in that tail.

How steep is exponential, really?

The curve is not a line, and that’s the entire lesson. Near room temperature a 10 K bump roughly doubles $k$ — and the curve keeps steepening as you go right.

You’ve probably heard the lab rule of thumb: “the rate roughly doubles for every 10 °C.” It gets repeated like folklore. It isn’t folklore — it’s a specific claim about a specific activation energy, and you can solve for it.

Take the ratio of $k$ at two temperatures. The $A$ cancels, which is the whole trick:

Now demand that $k_2/k_1 = 2$ when going from 300 K to 310 K, and solve for $E_a$:

Notice what that means for the $1/T$ in the exponent: the same 10 K rise buys you less at high temperature than at low. Going 300→310 K is a much bigger relative change in $1/T$ than 600→610 K. Temperature sensitivity itself fades as you get hotter.

The Arrhenius plot: how you actually measure $E_a$

An exponential is hard to read off a graph. So do what engineers always do with exponentials — take the log and straighten it out:

That’s $y = b + mx$ in disguise, with $y = \ln k$ and $x = 1/T$. Plot it and you get a straight line whose slope is $-E_a/R$ and whose intercept is $\ln A$.

Same reaction as the curve above, replotted. Measure the slope, multiply by $-R$, and the activation energy falls out — here recovering the 53.6 kJ/mol we started with.

Worked example: getting $E_a$ from two data points

You measure a rate constant at two temperatures:

• $k_1 = 0.0210\ \text{min}^{-1}$ at $T_1 = 298\ \text{K}$
• $k_2 = 0.0805\ \text{min}^{-1}$ at $T_2 = 318\ \text{K}$

Rearrange the two-point form to solve for $E_a$:

Run the numbers. The ratio is $k_2/k_1 = 0.0805/0.0210 = 3.833$, so $\ln(k_2/k_1) = 1.344$. And $\frac{1}{298} - \frac{1}{318} = 2.111\times10^{-4}$. Therefore:

And with $E_a$ in hand you can back out $A$ from either point: $A = k_1 e^{E_a/RT_1} = 0.0210 \times e^{52900/(8.314 \times 298)} \approx 4.0\times 10^{7}\ \text{min}^{-1}$.

Now you have the full temperature dependence and can predict $k$ at any temperature — which is exactly what you need to size a reactor that doesn’t run at the temperature you happened to measure at. A 20 K spread of data unlocks the whole curve.

Why you can’t just crank the heat

Here’s the trap. Temperature raises $k$ exponentially, so it looks like free conversion — want a smaller reactor? Just run hotter. And Part 1 said reactor volume scales with $1/(-r_A)$, so a hotter, faster reaction means a dramatically smaller vessel. All true.

But recall Part 2: your desired reaction is almost never the only one running. The side reaction has its own Arrhenius parameters. And selectivity between two competing paths depends on the ratio of their rate constants:

Everything hinges on the sign of $E_{a,D} - E_{a,U}$:

• If your desired reaction has the higher activation energy, heating up favors it. Selectivity improves with temperature. Run hot.
• If the side reaction has the higher activation energy, heating up feeds the junk faster than the product. You gain rate and lose selectivity. Run cool and accept a bigger reactor.

Add the safety dimension and it gets sharper still. An exothermic reaction releases heat, which raises $T$, which raises $k$ exponentially, which releases heat faster. That’s a positive feedback loop — thermal runaway — and it’s exponential on the way up. It’s precisely why Part 1 listed “taming a strong exotherm” as a reason to choose a CSTR: a big, well-mixed, diluted tank with lots of heat-transfer area is much harder to run away with than a PFR quietly growing a hot spot somewhere down the tube.

Where the series lands

Four articles, and they stack into one idea. The rate law ($-r_A = kC_A^n$) is the foundation, and it has exactly two knobs:

That’s reactor design. Pick where on the concentration axis you want to sit, pick where on the temperature axis you want to sit, and every volume, conversion, and yield number follows. The exponents come from data (Part 3); the reactor sets $C_A$; Arrhenius sets $k$. Everything else is algebra.