What a boundary surface diagram is
The orbital wave function $\psi$ for an electron in an atom has no physical meaning on its own; it is simply a mathematical function of the electron's coordinates. The quantity that carries physical meaning is $|\psi|^2$, the probability density, which by Max Born's interpretation gives the probability of finding the electron at a point. To picture an orbital's "shape" we therefore plot a surface built from this probability density rather than from $\psi$ itself.
A boundary surface diagram is a surface drawn in space on which $|\psi|^2$ has a constant value. In principle many such surfaces could be drawn for a single orbital. The one chosen to represent the orbital's shape is the surface that encloses the region within which the probability of finding the electron is very high — conventionally about 90%. This convention is not arbitrary: because $|\psi|^2$ retains some finite value, however small, at every finite distance from the nucleus, no rigid surface can ever enclose 100% of the probability.
A related way of picturing the same information is the charge-cloud diagram, in which the density of plotted dots in a region represents the electron probability density there. Where the dots are crowded the electron is most likely to be found; where they thin out the probability is low. The boundary surface diagram is essentially the clean outer envelope of such a cloud, drawn at the 90% contour. Both representations describe the most probable distribution of the electron in an orbital, and the shape they reveal is governed entirely by the azimuthal quantum number $l$ — it is $l$, not $n$, that decides whether the surface is a sphere, a dumb-bell or a cloverleaf.
"90% surface" is a deliberate cut-off, not a shortcoming of the diagram
Students sometimes mark "the boundary surface contains the whole electron" as correct. It does not. The electron's probability density extends to infinity, so the surface is drawn at the 90% probability contour purely so the picture is finite and useful.
Rule: the boundary surface is the contour of constant $|\psi|^2$ enclosing ~90% probability — never 100%.
Shape of s orbitals
For an s orbital the azimuthal quantum number is $l = 0$. The boundary surface of an s orbital is a sphere centred on the nucleus; in two dimensions it appears as a circle. This sphere encloses a region in which the probability of finding the electron is about 90%. All s orbitals — 1s, 2s, 3s and so on — are spherically symmetric, meaning the probability of finding the electron at a given distance is the same in every direction.
The size of an s orbital grows with the principal quantum number: $4s > 3s > 2s > 1s$. As $n$ increases, the electron is, on average, located further from the nucleus. The shape, however, stays spherical at every value of $n$ — what changes is the radius of the sphere and the appearance of internal radial nodes, discussed later.
Figure 1. Boundary surface of a 1s orbital. The surface is a sphere centred on the nucleus; the probability of finding the electron inside it is about 90%. All s orbitals share this spherical symmetry.
Shape of p orbitals
For a p orbital $l = 1$, and there are three permitted values of the magnetic quantum number ($m_l = -1, 0, +1$), so there are three p orbitals for any given $n \geq 2$. Unlike s orbitals, the boundary surfaces are not spherical. Each p orbital consists of two lobes lying on either side of a plane that passes through the nucleus — the familiar dumb-bell shape. The probability density falls to zero on that plane where the two lobes meet.
The three p orbitals are identical in size, shape and energy; they differ only in orientation. Their lobes lie along the $x$, $y$ and $z$ axes, giving them the labels $2p_x$, $2p_y$ and $2p_z$. Their axes are mutually perpendicular. It is worth noting that there is no simple one-to-one correspondence between the three $m_l$ values $(-1, 0, +1)$ and the three Cartesian directions; for NEET it is enough to know that three $m_l$ values produce three mutually perpendicular p orbitals. Like s orbitals, p orbitals grow in size and energy with $n$, so $4p > 3p > 2p$.
Figure 2. Boundary surfaces of the three 2p orbitals. Each is a two-lobed dumb-bell with the nucleus at the origin; the red dashed line marks the nodal plane through the nucleus on which the probability density is zero. The orbitals are identical except in orientation — along x, y and z.
The labels $l$, $m_l$ and "three p orbitals" come straight from the quantum numbers. Review Quantum Numbers (n, l, m, s) before fixing these shapes.
Shape of d orbitals
For $l = 2$ the orbital is a d orbital, and the smallest principal quantum number that can hold one is $n = 3$, because $l$ cannot exceed $n - 1$. There are five permitted $m_l$ values $(-2, -1, 0, +1, +2)$, so there are five d orbitals, designated $d_{xy}$, $d_{yz}$, $d_{xz}$, $d_{x^2-y^2}$ and $d_{z^2}$.
The shapes of the first four — $d_{xy}$, $d_{yz}$, $d_{xz}$ and $d_{x^2-y^2}$ — are similar to one another, each with four lobes arranged in a cloverleaf pattern. The fifth, $d_{z^2}$, has a distinct shape: two lobes directed along the $z$-axis together with a doughnut-shaped ring of electron density encircling the nucleus in the $xy$-plane. Despite the differing geometry, all five 3d orbitals are equal in energy (degenerate). Higher d orbitals (4d, 5d and so on) keep the same shapes but increase in size and energy.
Figure 3. Boundary surface of a representative d orbital, $d_{xy}$. Four lobes lie between the x and y axes; the two red dashed lines mark its two nodal planes (the xz- and yz-planes) passing through the nucleus, on which the probability density is zero — consistent with l = 2 giving two angular nodes.
Nodes, radial nodes and nodal planes
A node is a region where the probability density function $|\psi|^2$ reduces to zero — the electron is never found there. Nodes come in two kinds, and distinguishing them is a frequent source of errors. A radial node is a spherical surface at a particular distance $r$ from the nucleus on which the probability density is zero; for example, the 2s orbital has one radial node and 3s has two. An angular node, also called a nodal plane, is a plane passing through the nucleus on which the probability density is zero — the plane separating the lobes of a p orbital, for instance.
| Type of node | Geometry | Count formula | Depends on |
|---|---|---|---|
| Radial node | Spherical surface at distance r | n − l − 1 | Both n and l |
| Angular node (nodal plane) | Plane through the nucleus | l | l only |
| Total nodes | Radial + angular combined | n − 1 | n only |
The three formulas are linked: the total number of nodes is $(n-1)$, which is exactly the sum of the $l$ angular nodes and the $(n-l-1)$ radial nodes. The number of angular nodes equals $l$, so an s orbital ($l=0$) has no nodal plane, every p orbital ($l=1$) has exactly one nodal plane, and every d orbital ($l=2$) has two nodal planes. These relationships hold for any hydrogen-like orbital and are the basis of almost every node-counting question.
$$ \text{Total nodes} = (n-1) = \underbrace{l}_{\text{angular}} + \underbrace{(n-l-1)}_{\text{radial}} $$Nodal plane (angular) vs radial node — do not mix them
"Number of nodal planes" means angular nodes and equals $l$ alone — it never depends on $n$. So a 2p, a 3p and a 4p orbital each have exactly one nodal plane. The number that grows with $n$ is the radial-node count, $n-l-1$. A question asking for "nodal planes" wants $l$; a question asking for "radial nodes" wants $n-l-1$; "total nodes" wants $n-1$.
Rule: nodal planes = $l$ · radial nodes = $n-l-1$ · total nodes = $n-1$.
Worked node calculations
Applying the three formulas mechanically removes the guesswork. The following examples show the full breakdown for two commonly asked orbitals.
Find the number of radial nodes, angular nodes and total nodes in a 3p orbital.
For 3p: $n = 3$, $l = 1$.
Radial nodes $= n - l - 1 = 3 - 1 - 1 = 1$.
Angular nodes (nodal planes) $= l = 1$.
Total nodes $= n - 1 = 3 - 1 = 2$, which checks out as $1 + 1 = 2$. NCERT states the same: the number of radial nodes for 3p is one.
Find the number of radial nodes, angular nodes and total nodes in a 4d orbital.
For 4d: $n = 4$, $l = 2$.
Radial nodes $= n - l - 1 = 4 - 2 - 1 = 1$.
Angular nodes (nodal planes) $= l = 2$.
Total nodes $= n - 1 = 4 - 1 = 3$, consistent with $1 + 2 = 3$. So a 4d orbital has one radial node, two nodal planes and three nodes overall.
Orbital shape summary
The table consolidates the geometry, count and node data for the three orbital types tested at NEET. Note how the nodal-plane column tracks $l$ exactly while the shape is fixed by $l$ and unaffected by $n$.
| Orbital | l | No. of orbitals (2l+1) | Boundary-surface shape | Nodal planes (= l) |
|---|---|---|---|---|
| s | 0 | 1 | Spherical, centred on nucleus | 0 |
| p | 1 | 3 (px, py, pz) | Dumb-bell, two lobes along an axis | 1 |
| d | 2 | 5 (dxy, dyz, dxz, dx²–y², dz²) | Four-lobed cloverleaf; dz² has two lobes + ring | 2 |
Figure 4. Shape and nodal-plane count side by side: the spherical s orbital (0 nodal planes), the dumb-bell p orbital (1 nodal plane) and the cloverleaf d orbital (2 nodal planes). The number of nodal planes equals l in each case.
Shapes of atomic orbitals at a glance
- A boundary surface diagram is a constant-$|\psi|^2$ surface enclosing ~90% of the electron probability — never 100%, since $|\psi|^2$ is finite at all distances.
- s orbitals ($l=0$) are spherical and spherically symmetric; size grows as $4s>3s>2s>1s$.
- p orbitals ($l=1$) are dumb-bell shaped with two lobes; there are three — $p_x$, $p_y$, $p_z$ — differing only in orientation, each with one nodal plane.
- d orbitals ($l=2$) are five in number; four are cloverleaf-shaped and $d_{z^2}$ has two lobes plus a ring, yet all five are degenerate.
- Radial nodes $= n-l-1$; angular nodes (nodal planes) $= l$; total nodes $= n-1 = l + (n-l-1)$.