The bond order calculator above works the two routes to a bond order that ordinary chemistry uses. The first counts electrons in a molecular orbital diagram and takes half the difference between the bonding and antibonding populations. The second averages a bond across equivalent positions in a delocalised structure, which is how a nitrate ion ends up with three identical bonds that are each worth one and a third. Both are reported with the magnetic behaviour that the electron count implies, because that is the prediction most often needed alongside the number.
Arb Digital publishes free calculators for the calculations where the interpretation carries more weight than the arithmetic. Bond order is a case in point: dividing by two is not the hard part. Knowing which electrons to count, why a fractional answer is not a mistake, and why a bond order of zero is a genuine prediction rather than an error — that is where the value sits, and it is what most of this page is about.
What This Bond Order Calculator Does
In molecular orbital mode you supply the number of electrons in bonding orbitals, the number in antibonding orbitals, and how many are unpaired. It returns the bond order, the net bonding electrons, and whether the species is paramagnetic or diamagnetic. A species selector fills the boxes for the common homonuclear and heteronuclear diatomics if you want to check a result rather than derive one.
In resonance mode you supply the total number of bonds distributed across a set of equivalent positions in one Lewis structure and the number of those positions. It divides one into the other. For a nitrate ion that is four bonds — one double and two single — spread over three nitrogen-to-oxygen positions, giving one and a third. This is the same answer you get by drawing all three resonance structures and averaging one position across them, which is the longer route to the identical number.
The two modes are describing the same physical quantity from different starting points. Molecular orbital theory gives it directly from the electron distribution. The resonance average recovers it from Lewis structures, which cannot represent a delocalised bond in a single drawing and so need several drawings and an average to get there.
How to Use It
- Choose the method that matches what you have: an MO diagram, or a Lewis structure with equivalent bonds.
- Count the bonding electrons. These occupy the sigma and pi orbitals without an asterisk. Include core orbitals or exclude them, but do the same on both sides.
- Count the antibonding electrons, the ones in starred orbitals. Getting these wrong is the single commonest source of a wrong bond order.
- Count unpaired electrons from the diagram, remembering that degenerate orbitals fill singly before pairing.
- Read the bond order and the magnetism. A fractional value is a real answer, and a value of zero or below is the prediction that the species does not hold together.
The Formula and How It Is Calculated
The molecular orbital definition is bond order = (Nb − Na) / 2, where Nb is the number of electrons in bonding molecular orbitals and Na the number in antibonding ones. The division by two is there because a conventional single bond is two shared electrons, so the formula is counting bonds rather than electrons. The definition and the diagrams behind it are set out in the standard treatment of molecular orbital theory.
Working the default, dinitrogen has fourteen electrons. Ten occupy bonding orbitals and four occupy antibonding ones, so the bond order is (10 − 4)/2 = 3, and with no unpaired electrons it is diamagnetic. That is the triple bond a Lewis structure also predicts, and the agreement is reassuring rather than interesting. Dioxygen is where the two theories part company: sixteen electrons give ten bonding and six antibonding for a bond order of (10 − 6)/2 = 2, but two of those antibonding electrons sit singly in degenerate pi orbitals, so the molecule is paramagnetic.
The resonance average is simply bond order = total bonds / equivalent positions. A carbonate ion has four bonds distributed over three carbon-to-oxygen positions, giving 1.33. Benzene has nine bonds across the ring — six sigma and three pi — over six carbon-to-carbon positions, giving 1.5. In both cases the fractional answer is the point: no single Lewis structure can show it, and the experimental bond lengths confirm that all the bonds are identical rather than alternating.
Why Oxygen Is Paramagnetic and Lewis Structures Cannot Say So
Liquid oxygen sticks to a magnet. This is a demonstration rather than a subtlety, and it is the clearest experimental failure of Lewis theory in introductory chemistry. The Lewis structure of dioxygen shows a double bond and two lone pairs on each atom, with every electron paired, which predicts a diamagnetic molecule. The prediction is simply wrong.
Molecular orbital theory gets it right without any special pleading. Filling the orbitals in energy order puts the last two electrons into the two degenerate antibonding pi orbitals. Hund's rule places them one in each with parallel spins rather than pairing them in one, and two unpaired electrons make the molecule paramagnetic. The bond order still comes out as two, matching the double bond, so the theory reproduces the part Lewis got right while also explaining the part it got wrong.
This is why the unpaired electron count is a separate input on this page rather than something derived from the bond order. Bond order and magnetism are independent pieces of information from the same diagram. Nitrogen monoxide has a bond order of 2.5 and one unpaired electron; the peroxide ion has a bond order of one and none. Neither figure predicts the other.
Fractional Bond Orders Are Answers, Not Errors
A bond order of 1.5 or 2.5 looks like a rounding failure to anyone taught that bonds come in ones, twos and threes. It is not. A fractional bond order means the bonding electron density between two atoms does not correspond to a whole number of electron pairs, which happens in two distinct situations.
The first is an odd electron count. Nitrogen monoxide has fifteen electrons, so the difference between bonding and antibonding populations is odd and halving it gives a half. The molecule is a stable radical, and its bond order of 2.5 sits sensibly between the double bond of dioxygen and the triple bond of dinitrogen, which is exactly where its measured bond length falls.
The second is delocalisation. In a nitrate ion no single position holds a double bond; the pi system is spread over all three oxygen atoms equally. The three bonds are experimentally identical, and each is worth one and a third. Reporting them as one double and two singles would be a description of the drawing rather than of the ion. The same logic gives benzene its uniform 1.5 bonds and its uniform carbon-to-carbon distances, intermediate between a single and a double bond.
The one value that means something categorical rather than quantitative is zero. A bond order of zero says the antibonding population cancels the bonding population and there is no net bond, which is the prediction that the species does not exist as a stable molecule. Dihelium is the textbook case, and it is the correct answer rather than a failed calculation.
What Bond Order Predicts About Length and Strength
Across a series of related species, bond order tracks two measurable quantities well. Higher bond order means a shorter bond, because more shared electron density pulls the nuclei together, and it means a stronger bond, because more energy is needed to separate them. The oxygen series shows this cleanly: dioxygen at bond order two, superoxide at 1.5 and peroxide at one have progressively longer and weaker bonds, in that order.
The important qualifier is that this holds within a series, not across the periodic table. A carbon-to-carbon single bond and a hydrogen-to-fluorine single bond both have a bond order of one and have very different lengths and strengths, because the atoms are different sizes with different electronegativities. Bond order is one factor among several, and comparing it between unrelated pairs of atoms is meaningless. Where the comparison is about polarity rather than count, the electronegativity calculator is the right tool, and for ionic solids the lattice energy calculator covers the equivalent question.
Measured bond lengths and dissociation energies for real species, against which any predicted trend can be checked, are held in the NIST Chemistry WebBook. Checking a prediction against measurement is worth doing, because it exposes quickly whether a series is genuinely comparable.
Where the Simple Electron Count Breaks Down
The formula on this page assumes a two-centre picture: two atoms, a set of orbitals formed between them, and electrons that are either bonding or antibonding with respect to that pair. Several common situations do not fit.
Non-bonding orbitals are the mildest case. Electrons in an orbital that is neither bonding nor antibonding between the two atoms in question contribute nothing to the bond order and must be left out of both totals. Including them in the bonding count inflates the answer, and it is an easy mistake when reading a diagram quickly.
Heteronuclear diatomics need more care, because the two atoms contribute orbitals at different energies and the resulting molecular orbitals are not shared evenly. The bond order arithmetic still works, but the orbitals are no longer symmetric and their character is weighted towards one atom. Beyond diatomics, a polyatomic molecule has molecular orbitals spread over the whole framework rather than over one pair, so a bond order between a particular pair of atoms is no longer read off directly from a global electron count; it is calculated from the orbital coefficients, and computational chemistry offers several competing definitions that do not always agree.
Transition metal complexes are the furthest from the simple picture, with metal-to-metal bonding, delta bonds and bond orders as high as four or more that no counting scheme from a main-group diagram anticipates. If you are working with the electron configurations that underlie any of this, the electron configuration calculator and the effective nuclear charge calculator cover the atomic side.
Arb Digital publishes hundreds of free calculators across chemistry, maths, finance and marketing — no sign-up, no limits. If something you need is missing, tell us and we will look at building it.
Browse All Free Tools Suggest a ToolCommon Mistakes to Avoid
- Forgetting to halve the difference — the formula counts electron pairs, so the net electron count must be divided by two.
- Counting core electrons on one side only — they cancel exactly, so include them in both totals or in neither.
- Adding non-bonding electrons to the bonding total — lone pairs in orbitals that are neither bonding nor antibonding contribute nothing.
- Pairing electrons in degenerate orbitals too early — Hund's rule fills them singly first, which is what makes dioxygen paramagnetic.
- Treating a fractional result as an error — odd electron counts and delocalised systems genuinely give halves and thirds.
Related Free Tools From Arb Digital
Build the underlying electron picture with the electron configuration calculator and the effective nuclear charge calculator. For bond polarity rather than bond count, use the electronegativity calculator, and for ionic rather than covalent bonding the lattice energy calculator. On the reaction side, the chemical equation balancer and the net ionic equation calculator handle the equations, and the molar mass calculator turns any formula into g/mol. The full free online tools hub lists everything else.
Frequently Asked Questions
It is the number of chemical bonds between a pair of atoms. In molecular orbital theory it is half the difference between the number of bonding electrons and the number of antibonding electrons.
It makes no difference either way, because core orbitals contribute equally to the bonding and antibonding totals and cancel exactly. What matters is being consistent within one calculation rather than counting them on one side only.
That the antibonding electrons cancel the bonding electrons and there is no net bond. It is the prediction that the species does not exist as a stable molecule, which is why dihelium has never been isolated under ordinary conditions.
Either the species has an odd number of electrons, as nitrogen monoxide does, or the bonding is delocalised over several equivalent positions, as in a nitrate ion or benzene. Both give genuine fractional values.
It does not predict it. Magnetism depends on whether any electrons are unpaired, which is separate information from the bonding and antibonding totals. Dioxygen has a bond order of two and is paramagnetic because two electrons sit singly in degenerate orbitals.
Within a related series, yes: higher bond order means shorter and stronger. Across unrelated pairs of atoms it does not, because atomic size and electronegativity also set the length and the strength.
Divide the total number of bonds shown in one Lewis structure by the number of equivalent positions they are spread over. A nitrate ion has four bonds across three positions, so each bond order is one and a third.
This calculator is provided for education and general reference. It computes the standard molecular orbital and resonance definitions of bond order and is not a substitute for a full computational treatment of a real molecule.
