Orbital shapes — s (spherical), p (dumbbell), d (cloverleaf), f: Worked examples

Intuition What this page is for

The parent note gave you the rules: shape lives in the angular part $Y_{ℓ}^{m}$ , angular nodes $= ℓ$ , radial nodes $= n - ℓ - 1$ , total nodes $= n - 1$ . This page makes sure you never meet a case you have not already practised. We march through every kind of question node-counting can throw at you — every subshell, the degenerate ground state, the "wrong quantum-number" trap, the doughnut orbital, a bonding twist, and a limiting-case thought experiment.

Before we start, four words we will lean on:

Node = a place where the wavefunction $ψ$ is exactly zero, so the electron is never found there. Picture a still line on a vibrating guitar string — that motionless point is a node. In 3D the "still points" become whole surfaces.
We split those surfaces into radial nodes (thin spherical shells, like the layers of an onion where probability drops to zero) and angular nodes (flat planes or curved cones slicing through the nucleus).
The symbol $Y_{ℓ}^{m}$ has two labels riding on it. $ℓ$ (read "ell") is the shape number — it fixes how many angular nodes there are. $m$ (the magnetic quantum number) picks which orientation in space that shape takes — it is why there are separate $p_{x}, p_{y}, p_{z}$ rather than one $p$ . So $ℓ$ = "what shape", $m$ = "pointing which way". Full story in Quantum Numbers n, l, m, s.
$R_{n, ℓ} (r)$ (read "R-sub-n-ell of r") is the radial part of the orbital — a formula that tells you how the wavefunction's amplitude rises and falls purely as a function of the distance $r$ from the nucleus (ignoring direction). It is the other factor of $ψ = R_{n, ℓ} (r) Y_{ℓ}^{m} (θ, ϕ)$ . When $R_{n, ℓ} (r) = 0$ at some distance, you get a radial node (a spherical shell). Its explicit shape (a polynomial in $r$ times a decaying exponential $e^{- r / (something)}$ ) comes straight out of the Schrodinger Equation for Hydrogen; you can see its bumps plotted in the Radial Distribution Function note. We only need one fact about it here: it changes sign exactly at each radial node.

See Quantum Numbers n, l, m, s for where $n, ℓ, m$ come from, and the parent topic for the shapes themselves. (That parent title ends in "Hinglish" only because the vault keeps a Hindi–English explainer of the same topic — it is just the note's filename, nothing to do with f orbitals.)

The scenario matrix

Every question about orbital shape / nodes falls into one of these cells. The worked examples below are tagged with the cell(s) they cover so you can see the whole space is filled.

Cell	What makes it special	Covered by
A. $ℓ = 0$ (s)	zero angular nodes — the round case	Ex 1
B. $ℓ = 1$ (p)	one nodal plane	Ex 2
C. $ℓ = 2$ cloverleaf	two nodal planes, four lobes	Ex 3
D. $ℓ = 2$ the odd one $d_{z^{2}}$	nodes are cones, not planes	Ex 4
E. $ℓ = 3$ (f orbital)	three angular nodes — the extreme	Ex 5
F. Degenerate / smallest case ( $1 s$ )	zero of everything — the "empty" limit	Ex 6
G. Illegal / zero-degenerate input	$ℓ \geq n$ forbidden; catch the trap	Ex 7
H. Real-world word problem	count nodes from a spectroscopy label	Ex 8
I. Exam twist (sign & bonding)	the $+ / -$ lobe meaning	Ex 9
J. Limiting behaviour	what happens as $n \to \infty$ at fixed $ℓ$	Ex 10

Two formulas power the whole table — keep them in view:

Formula The node toolkit

$angular nodes = ℓ, radial nodes = n - ℓ - 1, total = n - 1.$ Why angular nodes $= ℓ$ ? The angular part $Y_{ℓ}^{m}$ is a polynomial of degree $ℓ$ in the direction cosines $x / r, y / r, z / r$ , so it can hit zero on at most $ℓ$ surfaces. A subtlety about $m$ (matters for $ℓ > 2$ ): the spherical harmonics $Y_{ℓ}^{m}$ are a set of $2 ℓ + 1$ normalised, mutually orthogonal functions (their squared-magnitudes integrate to $1$ over the sphere, and different ones never overlap — see Spherical Harmonics). Fixing $ℓ$ fixes the total number of angular nodes at $ℓ$ , but the value of $m$ decides how those $ℓ$ nodes are split between "planes/cones through the poles" and "planes containing the $z$ -axis". Concretely, $∣ m ∣$ nodal surfaces contain the $z$ -axis, and the remaining $ℓ - ∣ m ∣$ are cones/planes cutting across it. That is why $d_{z^{2}}$ ( $m = 0$ , both nodes are cones) looks utterly different from $d_{x y}$ (two axis-containing planes) even though both have $ℓ = 2$ and therefore both have exactly two angular nodes. For $f$ orbitals ( $ℓ = 3$ ) this $m$ -bookkeeping is what produces the several different named shapes. Why total nodes $= n - 1$ ? Here is the quick intuition: the number $n$ counts how many "half-waves" the electron's wavefunction packs between the nucleus and infinity — exactly like the $n$ -th vibration mode of a string has $n - 1$ still points between its two ends. One end (the nucleus region) and the far end (infinity) are not counted, so a mode labelled $n$ leaves $n - 1$ genuine zero-crossings in between. That count is fixed by $n$ alone. Splitting those $n - 1$ zeros into angular ( $ℓ$ ) and radial then forces radial $= (n - 1) - ℓ = n - ℓ - 1$ . (Full derivation in the parent note and Spherical Harmonics.)

Worked examples

Ex 1 — Cell A: the pure sphere ( $3 s$ )

Worked example Count and locate every node in a

$3 s$ orbital, and say what shape survives. Forecast: guess now — how many total nodes, and are any of them flat planes?

Read $n$ and $ℓ$ . The label " $3 s$ " means $n = 3$ and $s \Rightarrow ℓ = 0$ . Why this step? Every node count starts from these two numbers; nothing else matters.
Angular nodes $= ℓ = 0$ . Why? $Y_{0}^{0} = \frac{1}{4 π}$ is a constant — the same value in every direction — so it never reaches zero. No plane, no cone. That is exactly why the shape is a perfect ball.
Radial nodes $= n - ℓ - 1 = 3 - 0 - 1 = 2$ . Why? Total nodes must be $n - 1 = 2$ , and step 2 used none of them on angles, so all $2$ are spherical shells. At each such shell the radial part $R_{n, ℓ} (r)$ (defined in the intro — the distance-only factor of $ψ$ ) flips sign: positive amplitude inside, negative in the next layer, and so on — the shell is exactly where that sign change passes through zero.
Picture it (figure below): a solid centre, then a hollow spherical shell (node), a bright shell, another hollow shell (node), a faint outer shell. Onion, not clover.

Verify: angular $+$ radial $= 0 + 2 = 2 = n - 1$ ✔. Shape stays spherical because roundness is set by $ℓ$ , never by $n$ (see the parent's mistake box).

Worked example Figure 1 — what to look for

Figure — Orbital shapes — s (spherical), p (dumbbell), d (cloverleaf), f

Alt-text: concentric pastel rings for a 3s orbital. Key features: the three filled shells (lavender core, mint, butter) are regions of high probability; the two dashed coral circles are the radial nodes where

ψ = 0

. Notice there is no straight line or wedge cutting through the middle — that visual absence is the statement "zero angular nodes", i.e. perfectly round. Each dashed circle is a root of the radial polynomial

R_{3, 0} (r)

: the function that made the bright shells crosses zero there and returns with the opposite sign.

Ex 2 — Cell B: one nodal plane ( $2 p_{z}$ )

Worked example Where is the node of

$2 p_{z}$ , and which axis does the dumbbell lie on? Forecast: flat plane or curved cone? Which plane?

Numbers: $2 p_{z} \Rightarrow n = 2, ℓ = 1$ . (The subscript $z$ is the $m$ -label telling us this $p$ points along $z$ .)
Angular nodes $= ℓ = 1$ — one surface, and for a $p$ it is a plane. Why a plane? The angular part is $Y \propto cos θ$ , where $θ$ is measured down from the $z$ -axis. $cos θ$ is a first-degree (linear) shape, and a linear function vanishes on a flat surface, not a curved one.
Find where $cos θ = 0$ : $θ = 9 0^{\circ}$ . That is the whole $x y$ -plane (everything at right angles to $z$ ). Why $9 0^{\circ}$ ? $cos$ of a right angle is zero — the electron density is squeezed out of the equator.
Find the maxima: $cos θ = \pm 1$ at $θ = 0^{\circ}$ (straight up, $+ z$ ) and $θ = 18 0^{\circ}$ (straight down, $- z$ ). Two lobes on the $z$ -axis → dumbbell along $z$ .
Radial nodes $= n - ℓ - 1 = 2 - 1 - 1 = 0$ — no onion shells, just the two clean lobes.

Verify: total $= 1$ angular $+ 0$ radial $= 1 = n - 1 = 2 - 1$ ✔. The upper lobe carries $+$ sign of $ψ$ , lower carries $-$ (matters only for bonding — see Ex 9).

Worked example Figure 2 — what to look for

Alt-text: two vertical lobes of a p orbital split by a horizontal line. Key features: the lavender (+) lobe and coral (−) lobe sit on the

z

-axis; the thick mint horizontal line is the

x y

nodal plane at

θ = 9 0^{\circ}

. The two arrows mark the

cos θ = \pm 1

maxima. The opposite colours drive home that the sign of

ψ

flips across the plane — a fact that only matters for bonding, not charge.

Ex 3 — Cell C: two nodal planes ( $3 d_{x y}$ )

Worked example Sketch the node structure of

$3 d_{x y}$ (a cloverleaf d orbital). Forecast: how many lobes, and how many planes cut them apart?

Numbers: $3 d_{x y} \Rightarrow n = 3, ℓ = 2$ . (The " $x y$ " is the $m$ -label fixing the orientation between the $x$ and $y$ axes.)
Angular nodes $= ℓ = 2$ . For $d_{x y}$ the angular part $\propto x y / r^{2}$ . Why two planes here? The product $x y$ is zero whenever $x = 0$ or $y = 0$ . Those are two flat planes — the $y z$ -plane and the $x z$ -plane — crossing at right angles through the nucleus.
The two crossing planes cut space into four wedges → four lobes sitting between the axes (in the $x y$ -plane, pointing into the corners). This is the cloverleaf. Why does each new node double the lobe count? Start with one blob. The first nodal surface slices it in two (2 lobes). The second surface passes through both of those, slicing each again → $2 \times 2 = 4$ . Every extra node surface cuts through every existing lobe once, so the count multiplies by $2$ each time: $ℓ$ nodes → up to $2^{ℓ}$ lobes.
Radial nodes $= n - ℓ - 1 = 3 - 2 - 1 = 0$ — no extra shells for $3 d$ ; the lobes are solid.

Verify: $2$ angular $+ 0$ radial $= 2 = n - 1$ ✔. Four lobes, alternating sign $+, -, +, -$ around the ring, and $2^{ℓ} = 2^{2} = 4$ ✔.

Worked example Figure 3 — what to look for

Alt-text: four-lobed clover between the axes, split by the two axis lines. Key features: the two thick mint lines are the

x = 0

and

y = 0

nodal planes; the four lobes (

+, -, +, -

, lavender/coral) sit in the wedges between the axes. Reading round the ring you see the sign alternate at each plane crossing — the picture of

2^{2} = 4

lobes made by two cuts.

Ex 4 — Cell D: the doughnut orbital ( $3 d_{z^{2}}$ )

Worked example The odd d orbital: show that

$3 d_{z^{2}}$ still has exactly 2 angular nodes even though it looks nothing like a cloverleaf — and mark the sign of every region. Forecast: it has one big lobe up, one down, and a ring around the middle. Where can two nodes possibly be, and which parts are $+$ ?

Numbers: $n = 3, ℓ = 2 \Rightarrow$ angular nodes must be $2$ (the rule never lies). Here $m = 0$ , so by the $m$ -bookkeeping in the toolkit box both nodes are cones (none contains the $z$ -axis).
Angular part $\propto 3 cos^{2} θ - 1$ . Why this form? Unlike $d_{x y}$ , this one depends only on $θ$ (tilt from $z$ ), not on $ϕ$ (spin around $z$ ) — that is why it is symmetric like a doughnut. This function is exactly the associated Legendre polynomial $P_{2}^{0} (cos θ)$ ; its roots are the nodal cones.
Set it to zero: $3 cos^{2} θ - 1 = 0 \Rightarrow cos^{2} θ = \frac{1}{3} \Rightarrow cos θ = \pm \frac{1}{3}$ . Why solve this? These $θ$ values are exactly where $ψ = 0$ — the nodes.
Convert to an angle: $θ = arccos (\frac{1}{3}) \approx 54. 7^{\circ}$ and its mirror $\approx 125. 3^{\circ}$ . Each fixed $θ$ is a cone opening around the $z$ -axis. Two values of $θ$ → two cones → still two angular nodes, just cone-shaped instead of flat.
The signs (crucial for Ex 9): $3 cos^{2} θ - 1$ is positive where $cos^{2} θ > \frac{1}{3}$ , i.e. near the poles ( $θ < 54. 7^{\circ}$ or $θ > 125. 3^{\circ}$ ) → the top and bottom lobes are $+$ . It is negative in the waist band ( $54. 7^{\circ} < θ < 125. 3^{\circ}$ ) → the torus is $-$ . The sign flips as you cross each cone.

Verify: $cos^{- 1} (1/ 3) \approx 54.7 4^{\circ}$ , and $18 0^{\circ} - 54.7 4^{\circ} = 125.2 6^{\circ}$ ; two solutions → two cones → $2$ angular nodes ✔, matching $ℓ = 2$ . Sign pattern $+ / - / +$ from pole to waist to pole ✔.

Worked example Figure 4 — what to look for

Alt-text: a big top and bottom lobe on z with a small waist ring, plus dashed diagonal cone lines, coloured by sign. Key features: the dashed mint diagonals are the two nodal cones — and these lines are literally the roots of the Legendre polynomial

P_{2}^{0} (cos θ) = 3 cos^{2} θ - 1

: read

θ

off the diagonal and

3 cos^{2} θ - 1

is zero exactly there. The lavender ( $+$ ) lobes run up/down the

z

-axis (where the polynomial is positive); the pinched coral ( $-$ ) torus hugs the middle (where it is negative). Crossing a cone flips lavender↔coral — that is the sign change of

ψ

that Ex 9 needs for bonding.

Ex 5 — Cell E: the extreme ( $4 f_{z^{3}}$ )

Worked example Count nodes in a

$4 f$ orbital, and — picking a specific one, $4 f_{z^{3}}$ — say exactly how its three angular nodes are arranged and how many lobes result. Forecast: $f$ is the busiest common orbital. For the $z^{3}$ one, planes, cones, or a mix?

Numbers: $4 f \Rightarrow n = 4, ℓ = 3$ .
Angular nodes $= ℓ = 3$ (any $4 f$ ). Why 3? $Y_{3}^{m}$ is a degree-3 polynomial → up to three surfaces where it vanishes. But which surfaces depends on $m$ — so we must name the orbital to draw it.
Pick $4 f_{z^{3}}$ ( $m = 0$ ). Its angular part is $\propto 5 cos^{3} θ - 3 cos θ = cos θ (5 cos^{2} θ - 3)$ — the associated Legendre polynomial $P_{3}^{0} (cos θ)$ . Factor it to read off the roots:
- $cos θ = 0 \Rightarrow θ = 9 0^{\circ}$ : the $x y$ -plane (one flat node).
- $5 cos^{2} θ - 3 = 0 \Rightarrow cos θ = \pm 3/5 \Rightarrow θ \approx 39. 2^{\circ}, 140. 8^{\circ}$ : two cones. So the three angular nodes are one plane + two cones (not three planes). This is the honest picture the naive " $2^{ℓ}$ lobes" slogan hides. Why factor? A polynomial is zero exactly at its roots; factoring shows the roots, and each root is one nodal surface.
Lobe count for $4 f_{z^{3}}$ : the plane and two cones carve the $z$ -axis into two large axial lobes ( $+$ and $-$ ) plus two smaller belt-shaped lobes near the waist → six regions, alternating in sign — not eight. The " $2^{ℓ} = 8$ " figure is only an upper bound reached by orbitals whose nodes are three mutually-crossing planes (like $f_{x y z}$ ).
Radial nodes $= n - ℓ - 1 = 4 - 3 - 1 = 0$ — none; $4 f$ is the first $f$ level (you need $n \geq 4$ before $ℓ = 3$ is allowed).

Verify: $3$ angular $+ 0$ radial $= 3 = n - 1 = 4 - 1$ ✔. Roots of $P_{3}^{0}$ : $cos θ \in {0, \pm 3/5}$ giving $9 0^{\circ}, 39. 2^{\circ}, 140. 8^{\circ}$ ✔. Upper bound $2^{ℓ} = 2^{3} = 8$ lobes; $f_{z^{3}}$ realises $6$ .

Ex 6 — Cell F: the "empty" ground state ( $1 s$ )

Ex 7 — Cell G: the illegal input (catch the trap)

Ex 8 — Cell H: real-world word problem

Worked example A spectroscopist records an emission line from an electron sitting in the

$5 p$ state of an atom. For their line-shape model they need the total number of nodal surfaces of that orbital, split into planes and shells. Provide it. Forecast: more or fewer than 4 total surfaces?

Translate the label: " $5 p$ " → $n = 5, ℓ = 1$ . Why first? Physics words hide quantum numbers; extract them before computing.
Angular nodes $= ℓ = 1$ — one nodal plane (the $p$ dumbbell's single flat cut).
Radial nodes $= n - ℓ - 1 = 5 - 1 - 1 = 3$ — three onion-shell surfaces, and the radial part $R_{5, 1} (r)$ changes sign three times as you move outward (its polynomial factor has three positive roots).
Total $= 1 + 3 = 4$ , which must equal $n - 1 = 4$ . Physical meaning: the electron's probability cloud has four "quiet" surfaces it never occupies — one flat plane through the nucleus and three nested spheres.

Verify: $1 + 3 = 4 = n - 1 = 5 - 1$ ✔.

Ex 9 — Cell I: exam twist (the sign of a lobe)

Worked example A

$p_{z}$ orbital has a " $+$ " upper lobe and a " $-$ " lower lobe. During bond formation two such orbitals overlap $+$ lobe to $+$ lobe. State what the sign means and what this overlap produces — then contrast with the misread. Forecast: does " $+$ " mean positive electric charge?

What the sign is: the $+ / -$ is the ==sign of the wavefunction $ψ$ == in that lobe, not electric charge. (Across an angular node the sign flips because the angular part $Y_{ℓ}^{m}$ changes sign — exactly the pole→torus→pole flip we colour-coded in Ex 4; across a radial node it flips because $R_{n, ℓ}$ changes sign — either way it is a genuine $ψ$ sign change.) Why can't it be charge? Measurable electron density is $∣ ψ ∣^{2}$ , which squares away any sign and is always positive. Charge is never negative-then-positive across one electron's cloud.
Why the sign still matters: when two orbitals meet, their $ψ$ values add. Same sign meeting same sign → constructive (in-phase) overlap → density builds up between the nuclei → a bonding interaction.
Contrast the misread: if you thought " $+$ " meant positive charge, you'd predict repulsion — exactly wrong. The physics is wave interference, not electrostatics of the lobes.

Verify (conceptual): $∣ ψ ∣^{2} \geq 0$ everywhere regardless of the sign of $ψ$ , so lobe signs cannot be charges. See Bonding — Sigma and Pi Overlap.

Ex 10 — Cell J: limiting behaviour ( $n p$ as $n \to \infty$ )

Worked example Fix the shape at

$p$ ( $ℓ = 1$ ) and let $n$ grow: $2 p, 3 p, 4 p, \dots$ What stays constant, and what grows without bound? Forecast: does the dumbbell ever turn into a sphere for huge $n$ ?

Angular nodes $= ℓ = 1$ , always. Why constant? Shape is set by $ℓ$ , and we held $ℓ$ fixed. The single nodal plane never disappears → it stays a dumbbell for every $n$ .
Radial nodes $= n - ℓ - 1 = n - 2$ grows. For $2 p$ : $0$ . For $3 p$ : $1$ . For $10 p$ : $8$ . As $n \to \infty$ the count $\to \infty$ — and the radial part $R_{n, 1} (r)$ flips sign that many times. Why? Each rise in $n$ adds one more onion shell of zero probability while keeping the two-lobed silhouette.
Physical limit: the orbital gets ever larger and more layered but never rounds off; the electron becomes ever less bound (energy $\to 0$ ), approaching ionisation.

Verify: radial nodes for $10 p = 10 - 1 - 1 = 8$ , and it equals $n - 2$ ; total $= 1 + (n - 2) = n - 1$ ✔ for all $n$ .

Active recall

Recall Which cell does each question hit?

" $4 d$ node count" ::: Cell C — $ℓ = 2$ , radial $= 4 - 2 - 1 = 1$ , angular $= 2$ , total $3$ . "Why is $6 s$ round?" ::: Cell A — $ℓ = 0$ , angular part constant. "Does $d_{z^{2}}$ break the node rule?" ::: Cell D — no; its 2 nodes are cones not planes. "How are $4 f_{z^{3}}$ 's three nodes arranged?" ::: Cell E — one plane ( $9 0^{\circ}$ ) + two cones ( $39. 2^{\circ}, 140. 8^{\circ}$ ), giving 6 lobes. "Is $3 f$ allowed?" ::: Cell G — no; $ℓ = 3$ needs $n \geq 4$ . " $+$ lobe = positive charge?" ::: Cell I — no; sign of $ψ$ , for overlap only. "What does $m$ decide?" ::: the orientation of the shape and how the $ℓ$ angular nodes split into planes vs cones. "Where does the radial part $R_{n, ℓ}$ change sign?" ::: at each radial node.

Connections

Parent: Orbital shapes — the shapes these counts describe.
Quantum Numbers n, l, m, s — where $n, ℓ, m$ (and the $ℓ < n$ rule) come from.
Radial Distribution Function — visualising $R_{n, ℓ} (r)$ and the shells / nodes where it changes sign.
Spherical Harmonics — why angular nodes $= ℓ$ , and how $m$ splits them into planes vs cones.
Schrodinger Equation for Hydrogen — the $R \cdot Y$ split and the explicit form of $R_{n, ℓ}$ .
Bonding — Sigma and Pi Overlap — the lobe-sign twist in Ex 9.
Heisenberg Uncertainty Principle — why these are clouds, not paths.