18.1 Alpha-Particle Scattering and Rutherford's Model
Fire alpha particles at a gold leaf a few hundred atoms thick: almost all pass straight through, but about one in eight thousand bounces back. Only a tiny, massive, positively charged nucleus can do that — and energy conservation tells us how close an alpha particle gets to it, $r_0=2kZe^2/K$.
By 1900 physicists knew that atoms contain electrons — light, negatively charged particles that can be pulled out of any material — and that an atom as a whole is neutral. So every atom must also hold positive charge, carrying almost all of its mass. But where is that positive charge? Is it spread through the whole atom like a jelly, or gathered somewhere? You cannot look inside an atom with a microscope; it is about $10^{-10}\ \text{m}$ across, far smaller than the wavelength of light. What you can do is throw something at it and watch how it bounces.
That is exactly what Hans Geiger and Ernest Marsden did in 1909, under Ernest Rutherford in Manchester, using fast alpha particles from a radioactive source as the projectiles and a thin gold leaf as the target. The results forced Rutherford to a startling picture: an atom is almost entirely empty space, with all its positive charge and nearly all its mass packed into a nucleus about ten thousand times smaller than the atom. In this section we follow that argument, use energy conservation to find how close an alpha particle can get to a nucleus — a JEE Main favourite — and see why Rutherford's atom, for all its success, could not be the whole story.
The Geiger–Marsden experiment
The model being tested. J. J. Thomson, who discovered the electron, had proposed that an atom is a sphere of radius about $10^{-10}\ \text{m}$ filled uniformly with positive charge, with the electrons embedded in it like seeds in a watermelon (the "plum-pudding" model). In such an atom the positive charge is spread thinly, so the electric field anywhere inside or near it is modest. A fast, heavy projectile should go through such atoms with only tiny deflections.
The projectiles. An alpha particle is a helium nucleus: charge $+2e$, mass about $4\ \text{u}=6.64\times10^{-27}\ \text{kg}$ — roughly $7300$ times the mass of an electron. Radioactive sources such as polonium emit alpha particles with kinetic energies of a few MeV ($1\ \text{MeV}=1.6\times10^{-13}\ \text{J}$), so a $5.0\ \text{MeV}$ alpha particle moves at about $1.6\times10^7\ \text{m/s}$. Because it is so heavy and so fast, the electrons of the target atoms cannot deflect it noticeably — much as a cricket ball is not deflected by a swarm of dust. Any large deflection must come from the positive, massive part of the atom.
The set-up. A narrow beam of alpha particles, selected by a lead block with a slit (a collimator), falls on a gold foil only about $2\times10^{-7}\ \text{m}$ thick — a few hundred atoms. Gold was used because it can be beaten into very thin leaf, so each alpha particle meets essentially one atom that matters, and because its nucleus carries a large charge ($Z=79$). The scattered alpha particles strike a small screen coated with zinc sulphide, which gives a tiny flash of light (a scintillation) for each hit; an observer counts the flashes through a microscope. The screen and microscope swing around the foil, so the number of particles scattered through each scattering angle $\theta$ (the angle between the incoming and outgoing directions) can be counted. The whole apparatus is in a vacuum, so the alpha particles are not stopped by air.
In the Geiger–Marsden experiment a collimated beam of alpha particles of a few MeV is directed at a very thin gold foil in a vacuum, and the number of particles scattered through each angle $\theta$ is counted with a movable zinc-sulphide screen. The observations were:
- Most alpha particles pass straight through the foil or are deflected by only a degree or so.
- A few are deflected through large angles; about one in $8000$ is turned back through more than $90^\circ$, and some through nearly $180^\circ$.
- The number scattered falls off very steeply as the angle increases.
Rutherford famously said that the back-scattering was "as if you fired a 15-inch shell at a piece of tissue paper and it came back and hit you". In Thomson's atom no such thing could happen, as the first example shows: the largest force a spread-out atom can exert lasts too short a time to turn the particle by more than a few hundredths of a degree, and the largest electric potential energy it can give the particle is a few keV — a thousand times too small to stop a $5\ \text{MeV}$ projectile. Even many small deflections in a row, each as likely to the left as to the right, cannot add up to a reversal often enough.
In the Geiger–Marsden experiment, the observation that most alpha particles go straight through the gold foil shows that
Rutherford's nuclear model and the impact parameter
Rutherford's insight was that a large deflection in a single encounter needs a very strong force, and a very strong electric force needs the charge to be concentrated in a very small region that the alpha particle can approach closely. Because the alpha particle is turned back rather than knocking the target forward, that region must also hold most of the atom's mass.
In the nuclear model of the atom:
- all the positive charge $+Ze$ and almost all the mass of the atom are concentrated in a tiny central nucleus, of radius about $10^{-15}$–$10^{-14}\ \text{m}$ (a few femtometres);
- $Z$ electrons, making the atom neutral, move around the nucleus at distances of about $10^{-10}\ \text{m}$, held by the electric attraction of the nucleus, as planets are held by the Sun — hence the name "planetary model";
- the rest of the atom is empty space.
$Z$ is the atomic number: $1$ for hydrogen, $79$ for gold.
The scattering of an alpha particle by a nucleus is then a two-body problem with a known force: the Coulomb repulsion $F=\dfrac{k(2e)(Ze)}{r^2}$. The nucleus of gold is about $50$ times heavier than the alpha particle, so for now we treat it as fixed. What decides how much a particular alpha particle is deflected is how squarely it is aimed at the nucleus.
The impact parameter $b$ of an alpha particle is the perpendicular distance of its initial line of motion (far from the nucleus) from the centre of the nucleus. It is the distance by which it would miss the nucleus if there were no force.
- $b=0$ (a head-on approach): the particle slows down, stops momentarily and comes straight back — $\theta=180^\circ$.
- Small $b$: strong force for a long time — large scattering angle.
- Large $b$: the particle stays far from the nucleus — small scattering angle; for $b$ much larger than the nuclear size, $\theta\approx0$.
So the scattering angle decreases as the impact parameter increases. Each path is a hyperbola with the nucleus at its outer focus.
In a beam the particles have impact parameters scattered at random, and only the very few aimed within a few tens of femtometres of some nucleus are deflected strongly. That is why large-angle scattering is so rare, and why the number counted falls so steeply with angle. Rutherford worked out the expected number at each angle from the Coulomb force on a point nucleus, and Geiger and Marsden's counts matched it over a huge range — confirming both the inverse-square force and the tiny size of the nucleus. Interactive 1 lets you aim the particles yourself.
In Interactive 1 choose gold and $K=5.0\ \text{MeV}$. With $b=0$ the particle comes straight back ($\theta=180^\circ$). Predict the scattering angle when you raise the impact parameter to $b=50\ \text{fm}$.
Not in the JEE Main 2026 list (which names only the experiment and the model) nor in JEE Advanced 2026. This box is for interest only; nothing here is used in the exercises, the Exam corner or the quiz.
Solving Newton's second law with the Coulomb force gives, for a fixed nucleus, $b=\dfrac{r_0}{2}\cot\dfrac{\theta}{2}$, where $r_0=\dfrac{2kZe^2}{K}$ is the head-on distance of the next subsection. The closest distance on a path with impact parameter $b$ is $\dfrac{r_0}{2}\left(1+\dfrac{1}{\sin(\theta/2)}\right)$. Particles aimed within a ring of radius $b$ around a nucleus are scattered through more than $\theta$; the area of the ring, $\pi b^2$, and the number of nuclei per unit area of foil give the fraction scattered beyond $\theta$, and from it Rutherford's prediction that the number scattered per unit solid angle varies as $1/\sin^4(\theta/2)$ — the "very steep fall" that Geiger and Marsden confirmed. Interactive 1 uses exactly these relations for its readouts.
Two alpha particles of the same energy approach the same gold nucleus with impact parameters $b_1$ and $b_2$, where $b_1\lt b_2$. Then
Distance of closest approach
Consider the simplest case: an alpha particle of kinetic energy $K$ aimed head-on ($b=0$) at a nucleus of charge $Ze$, which we take to be fixed (very heavy). As it approaches, the Coulomb repulsion slows it down; its kinetic energy turns into electric potential energy. At some distance $r_0$ it stops for an instant and then is pushed straight back. Since the electric force is conservative, mechanical energy is conserved: far away the energy is all kinetic ($U\to0$), and at the turning point it is all potential.
A particle of charge $q$ and kinetic energy $K$ (far away) moving straight towards a fixed nucleus of charge $Ze$ stops momentarily at the distance of closest approach $r_0$, given by
$$K=\frac{kqZe}{r_0},\qquad\text{so}\qquad r_0=\frac{kqZe}{K}.$$For an alpha particle ($q=2e$, mass $m$, speed $v$): $\ r_0=\dfrac{2kZe^2}{K}=\dfrac{4kZe^2}{mv^2}$, with $k=\dfrac{1}{4\pi\varepsilon_0}$. This assumes (i) the nucleus does not move (it is much heavier than the projectile, or held fixed), (ii) only the Coulomb force acts — the particle stays outside the nucleus — and (iii) the motion is non-relativistic.
A number worth remembering. With $k=9.0\times10^9\ \text{N m}^2\text{C}^{-2}$ and $e=1.6\times10^{-19}\ \text{C}$, $ke^2=2.304\times10^{-28}\ \text{J m}$. Dividing by $1.6\times10^{-13}\ \text{J/MeV}$ and by $10^{-15}\ \text{m/fm}$:
$$ke^2=1.44\ \text{MeV fm}\ \ (=1.44\ \text{eV nm}).$$So for an alpha particle, $r_0\ (\text{fm})=\dfrac{2\times1.44\times Z}{K\ (\text{MeV})}$. A $5.0\ \text{MeV}$ alpha particle and a gold nucleus give $r_0=\dfrac{2\times1.44\times79}{5.0}=45.5\ \text{fm}$. Since the particle came back without touching the nucleus, the gold nucleus must be smaller than this — Rutherford's first estimate of nuclear size. (We now know its radius is about $7\ \text{fm}$.)
Read the formula as a set of proportionalities, which is how JEE Main uses it: $r_0\propto\dfrac{qZ}{K}$. Doubling the kinetic energy halves $r_0$; doubling the speed quarters it ($K\propto v^2$); for the same momentum $p$, $K=p^2/2m$ and $r_0\propto qm$.
- Write the projectile's charge $q$ ($2e$ for an alpha particle, $e$ for a proton or deuteron) and the target's charge $Ze$.
- Fixed (heavy) target: all the kinetic energy becomes potential energy, $K=\dfrac{kqZe}{r_0}$. Work in MeV and fm with $ke^2=1.44\ \text{MeV fm}$.
- If the data give speed or momentum, convert first: $K=\tfrac12mv^2=\dfrac{p^2}{2m}$; for a particle accelerated from rest through a potential difference $V$, $K=qV$.
- Free target (light, unattached nucleus): at closest approach the two move with the same velocity. Use momentum conservation for that velocity; only the kinetic energy relative to the centre of mass, $K\,\dfrac{M}{M+m}$, turns into potential energy.
- Check: is $r_0$ larger than the sum of the nuclear radii? If not, nuclear forces act and the Coulomb picture fails.
In Interactive 2 choose gold, the nuclear model and $K=5.0\ \text{MeV}$; the turning point is at $45.5\ \text{fm}$. Predict the turning point when you double the kinetic energy to $10\ \text{MeV}$.
An alpha particle fired head-on at a fixed nucleus has distance of closest approach $r_0$. If its speed is doubled, the distance of closest approach becomes
Shortcomings of Rutherford's model
Rutherford's nucleus survived every test, and it is still part of our picture of the atom. The trouble lies with the electrons. In his model they must move around the nucleus — if they stood still they would fall straight in — and the simplest motion is a circle, with the Coulomb attraction providing the centripetal force.
For an electron (mass $m$, charge $-e$) moving uniformly in a circle of radius $r$ around a fixed nucleus of charge $Ze$, $\dfrac{mv^2}{r}=\dfrac{kZe^2}{r^2}$, so
$$v=\sqrt{\frac{kZe^2}{mr}},\qquad K=\frac{kZe^2}{2r},\qquad U=-\frac{kZe^2}{r},\qquad E=K+U=-\frac{kZe^2}{2r}.$$The total energy is negative (the electron is bound), $K=-E$ and $U=2E$. A smaller orbit has a lower (more negative) energy and a higher frequency of revolution $f=v/2\pi r$.
For hydrogen with $r=0.53\times10^{-10}\ \text{m}$ this gives $v=2.19\times10^6\ \text{m/s}$, $f=6.6\times10^{15}\ \text{Hz}$ and $E=-13.6\ \text{eV}$ — the right size for the energy needed to ionise hydrogen. But it leads to two serious contradictions:
- Atoms would not be stable. Classical electromagnetism says that an accelerated charge radiates electromagnetic waves. An electron in a circular orbit is always accelerating (centripetally), so it should continuously lose energy. Since $E=-kZe^2/2r$, losing energy means a smaller $r$: the electron should spiral into the nucleus — a calculation gives about $10^{-11}\ \text{s}$ for hydrogen. Real atoms last for ever.
- The spectrum would be continuous. The radiation would have the frequency of revolution, which rises steadily as the electron spirals in. A collection of atoms would therefore emit a continuous range of frequencies. In fact a hot gas of one element emits only certain sharp wavelengths — a line spectrum characteristic of that element (hydrogen's red line at $656\ \text{nm}$, for instance).
The nucleus was right; classical mechanics applied to the electrons was wrong. In the next section Niels Bohr keeps Rutherford's nucleus and repairs the electrons with quantum ideas — and gets hydrogen's spectral lines exactly right.
1. Charge of the alpha particle. It is $+2e$, so $r_0=\dfrac{k(2e)(Ze)}{K}=\dfrac{2kZe^2}{K}$. Using $kZe^2/K$ (as for a proton) halves the answer.
2. Speed versus energy. $r_0\propto1/K\propto1/v^2$: doubling the speed gives $r_0/4$, doubling the energy gives $r_0/2$. For equal momenta $r_0\propto qm$; for the same accelerating voltage $r_0$ is the same for every particle.
3. Units. Use $ke^2=1.44\ \text{MeV fm}$ with $K$ in MeV to get $r_0$ in fm; if you work in SI, convert MeV to joules ($\times1.6\times10^{-13}$). Answers of nuclear size are $10^{-15}$–$10^{-13}\ \text{m}$.
4. Free targets. The fixed-nucleus formula assumes the target cannot move. For a light free target, at closest approach both particles share a common velocity; only $K\,\dfrac{M}{M+m}$ becomes potential energy, so they do not get as close.
5. What the observations show. "Most go straight through" $\Rightarrow$ the atom is mostly empty space; "a few bounce back" $\Rightarrow$ the positive charge and the mass are concentrated in a tiny nucleus. Neither observation tells you how the electrons are arranged.
According to classical electromagnetism, an electron revolving round the nucleus in Rutherford's atom should
Exercises
Take $k=9.0\times10^9\ \text{N m}^2\text{C}^{-2}$, $e=1.6\times10^{-19}\ \text{C}$, $ke^2=1.44\ \text{MeV fm}$, $1\ \text{MeV}=1.6\times10^{-13}\ \text{J}$, the mass of an alpha particle $6.64\times10^{-27}\ \text{kg}$ and of an electron $9.1\times10^{-31}\ \text{kg}$. Treat target nuclei as fixed unless told otherwise.
- 1 Give two reasons why a very thin gold foil was used in the scattering experiment, and one reason why the apparatus was evacuated.
- 2 State the three main observations of the experiment and what Rutherford concluded from each.
- 3 In Thomson's model a gold atom is a sphere of radius $1.0\times10^{-10}\ \text{m}$ holding a uniform charge $+79e$. Find the electric field at its surface.
- 4 Find the electric field at the surface of a gold nucleus of radius $7.0\ \text{fm}$ (charge $+79e$).
- 5 An atom of radius $1.0\times10^{-10}\ \text{m}$ has a nucleus of radius $1.0\times10^{-14}\ \text{m}$. What fraction of the atom's volume does the nucleus occupy?
- 6 Explain why the electrons of the gold atoms cannot cause the large-angle scattering of alpha particles.
- 7 Three alpha particles of equal energy approach a nucleus with impact parameters $10\ \text{fm}$, $50\ \text{fm}$ and $100\ \text{fm}$. Arrange them in order of increasing scattering angle. Which impact parameter gives $\theta=180^\circ$?
- 8 An alpha particle makes a head-on elastic collision with a free electron at rest. Find the fraction $\Delta v/v$ by which the alpha particle's speed falls.
- 1 Find the distance of closest approach of a $4.0\ \text{MeV}$ alpha particle fired head-on at a gold nucleus ($Z=79$).
- 2 Repeat for an $8.0\ \text{MeV}$ alpha particle and a silver nucleus ($Z=47$).
- 3 Find the distance of closest approach of a $2.0\ \text{MeV}$ proton fired head-on at a gold nucleus.
- 4 What kinetic energy must an alpha particle have to come within $20\ \text{fm}$ of the centre of a gold nucleus?
- 5 A $6.0\ \text{MeV}$ alpha particle fired head-on at a nucleus comes to rest momentarily $24\ \text{fm}$ from its centre. Find the atomic number of the nucleus.
- 6 Find the electric potential energy, in MeV and in joules, of an alpha particle $30\ \text{fm}$ from the centre of a gold nucleus.
- 7 A $5.0\ \text{MeV}$ alpha particle approaches a gold nucleus head-on. Find its kinetic energy when it is $100\ \text{fm}$ from the centre.
- 8 Find the speed of a $5.0\ \text{MeV}$ alpha particle.
- 1 The kinetic energy of an alpha particle aimed head-on at a nucleus is doubled. By what factor does its distance of closest approach change?
- 2 Its speed is doubled instead. By what factor does the distance change?
- 3 A proton and an alpha particle with the same kinetic energy are fired head-on at the same nucleus. Find $r_p/r_\alpha$.
- 4 Find $r_p/r_\alpha$ if they have the same speed.
- 5 Find $r_p/r_\alpha$ if they have the same momentum.
- 6 A deuteron (charge $e$, mass $2\ \text{u}$) and an alpha particle are each accelerated from rest through the same potential difference and fired head-on at the same nucleus. Find $r_d/r_\alpha$.
- 7 Sketch the potential energy of an alpha particle against its distance from a nucleus, add the line of its total energy, and mark the distance of closest approach. Mark the region the particle can never enter.
- 8 A $7.0\ \text{MeV}$ alpha particle comes within $32.5\ \text{fm}$ of a gold nucleus. What kinetic energy would it need to come equally close to a copper nucleus ($Z=29$)?
- 1 A $4.0\ \text{MeV}$ alpha particle (mass $4\ \text{u}$) moves head-on towards a free oxygen-16 nucleus ($Z=8$, mass $16\ \text{u}$) at rest. Find the distance of closest approach.
- 2 Two alpha particles, each of kinetic energy $0.50\ \text{MeV}$, move directly towards each other. Find their least separation.
- 3 Find the force on a $5.0\ \text{MeV}$ alpha particle at its closest approach to a gold nucleus (head-on).
- 4 In Rutherford's model of hydrogen the electron moves in a circle of radius $0.53\times10^{-10}\ \text{m}$. Find its speed.
- 5 Find the frequency with which it revolves — the frequency classical physics says it should radiate.
- 6 Find its total energy in eV, and its kinetic and potential energies.
- 7 What is the least kinetic energy a proton needs, fired head-on, to reach the surface of a gold nucleus of radius $7.0\ \text{fm}$? (Treat the proton as a point.)
- 8 Explain why Rutherford's model predicts that atoms collapse, and why that is in conflict with the line spectrum of hydrogen.
E1. JEE Main · MCQ An alpha particle fired head-on at a fixed nucleus has distance of closest approach $d$. A proton moving with twice the speed of the alpha particle is fired head-on at the same nucleus. The proton's distance of closest approach is (A) $d$ (B) $d/2$ (C) $d/8$ (D) $d/4$
E2. JEE Main · Numerical An alpha particle of kinetic energy $9.1\ \text{MeV}$ is fired head-on at a gold nucleus ($Z=79$), which may be taken as fixed. Find its distance of closest approach in femtometres. Take $\dfrac{e^2}{4\pi\varepsilon_0}=1.44\ \text{MeV fm}$. Give the answer to the nearest integer.
E3. JEE Main · MCQ In the alpha-particle scattering experiment, the fact that a very small fraction of the alpha particles are deflected through more than $90^\circ$ led to the conclusion that (A) the positive charge and most of the mass of an atom are concentrated in a very small region (B) the electrons of the atom are distributed uniformly through it (C) the atom contains negatively charged particles of very small mass (D) the positive charge of the atom is spread uniformly through its whole volume
E4. JEE Main · Numerical An alpha particle of kinetic energy $E$, fired head-on at a fixed nucleus X, comes to rest momentarily at $36\ \text{fm}$ from its centre. A proton of kinetic energy $E/2$ is now fired head-on at the same nucleus. Find its distance of closest approach in fm. Give the answer to the nearest integer.
E5. JEE Main · MCQ An alpha particle of kinetic energy $K$ moves head-on towards another alpha particle that is initially at rest and free to move. Ignoring all but the electric force, their least separation is (with $k=1/4\pi\varepsilon_0$) (A) $\dfrac{4ke^2}{K}$ (B) $\dfrac{2ke^2}{K}$ (C) $\dfrac{8ke^2}{K}$ (D) $\dfrac{16ke^2}{K}$
E6. JEE Main · Numerical In Rutherford's model of the hydrogen atom an electron moves in a circular orbit of radius $0.53\times10^{-10}\ \text{m}$ around a proton. Find the magnitude of its electric potential energy in eV ($k=9.0\times10^9\ \text{N m}^2\text{C}^{-2}$, $e=1.6\times10^{-19}\ \text{C}$). Give the answer to the nearest integer.
Research question. Can you find the size of an object you cannot see, only by throwing things at it and counting how many are deflected — the way Rutherford "saw" the nucleus?
Why it matters. Almost everything we know about the inside of matter comes from scattering experiments: X-ray crystallography revealed the structure of DNA, electron scattering at Stanford measured the size of the proton, the Large Hadron Collider at CERN finds new particles from the debris of collisions, and medical CT scanners, airport baggage scanners and seismologists probing the Earth's core all reconstruct hidden objects from how waves or particles are scattered. The logic is the same as in the Geiger–Marsden experiment: the fraction of projectiles that are deflected tells you how much of the target area is "solid".
What to do.
- Ask a partner to tape a few identical hidden "nuclei" (coins or bottle caps) under a large sheet of cardboard raised on books, without telling you how many or where they are, and to note their total number $N$ and diameter $d$.
- Roll marbles at random positions across the full width $W$ of the board, from one edge, aiming without looking. Record for each roll whether the marble comes out deflected or straight. Do at least $200$ rolls.
- A marble of diameter $d_m$ is deflected if its centre passes within $\tfrac12(d+d_m)$ of a hidden centre, so the fraction deflected should be about $f=\dfrac{N(d+d_m)}{W}$. Use your count of $f$ to estimate $d$ (your partner tells you $N$).
- Compare with the true $d$. How does the uncertainty change with the number of rolls?
- Make a table: what in your model plays the part of the alpha particle, the nucleus, the gold foil, the impact parameter and the zinc-sulphide screen?
Tools: Large cardboard sheet, books or blocks, coins or caps, tape, a dozen marbles, a ruler, a tally sheet (or a spreadsheet to compute $f$ and its uncertainty $\sqrt{f(1-f)/n}$).
Hand in: Your tally, the estimate of $d$ with its uncertainty, the comparison table, and a paragraph explaining why Rutherford needed so many counts to see the rare large-angle events.
Going further. Real alpha particles are deflected by a long-range force, not by touching. Write a short spreadsheet that steps an alpha particle past a nucleus under Coulomb's law (as Interactive 1 does) and record the scattering angle for impact parameters from $0$ to $100\ \text{fm}$. How does the angle depend on $b$?
🧪 Self-check quiz
Q1. Gold was chosen as the target in the alpha-particle scattering experiment mainly because
Q2. The distance of closest approach of an alpha particle of kinetic energy $K$ fired head-on at a fixed nucleus of atomic number $Z$ is ($k=1/4\pi\varepsilon_0$)
Q3. An alpha particle is scattered through $180^\circ$ by a nucleus. Its impact parameter was
Q4. Rutherford's nuclear model of the atom could not explain
Q5. An alpha particle and a proton with the same kinetic energy are fired head-on at the same fixed nucleus. The ratio of the proton's distance of closest approach to the alpha particle's is