Covalent bonding is best understood as an electrostatic attraction created by shared electron density, not as atoms simply touching. The orbital model explains why bonds have direction and strength, while coordinate bonding shows that a shared pair can originate from one atom without becoming a different kind of bond.

When atomic orbitals overlap, electron density lies between two nuclei. Attraction of both positive nuclei to this shared negative density holds the atoms together. A covalent bond is directional because the electron density is concentrated in a particular region of space.

A single bond is normally one $\sigma$ bond. A double bond contains one $\sigma$ and one $\pi$ bond; a triple bond contains one $\sigma$ and two $\pi$ bonds. The weaker, exposed $\pi$ component makes multiple bonds especially important in organic reactivity.

A covalent bond is a shared pair of electrons. In a coordinate (dative) bond, both electrons in the newly shared pair originate from one donor species and enter an empty acceptor orbital. Once formed, the bond is not automatically weaker or chemically distinguishable solely because of its origin. Lewis structures help account for valence electrons and formal charges, but multiple valid resonance contributors may be needed where electrons are delocalised.

Atoms share electrons to achieve a more stable outer-shell arrangement. Hydrogen requires two electrons in its first shell; many second-period atoms form enough shared pairs to reach eight outer-shell electrons. This is a useful pattern, not an excuse to ignore supplied charges or known bonding.

Dot-and-cross diagrams track the origin of electrons. Use different marks for different atoms, show all outer electrons, and put brackets and the charge around ions. A displayed formula shows every bond but normally omits lone pairs; choose the representation requested by the question.

For the same pair of atoms, a higher bond order usually means more electron density between the nuclei, a shorter bond and a larger bond enthalpy. Thus a carbon-carbon triple bond is shorter and stronger than a carbon-carbon double bond, which is shorter and stronger than a carbon-carbon single bond.

These comparisons are valid only when the atoms being compared are the same. A bond's polarity and environment also influence its length and strength, so avoid treating one numerical bond-energy value as universal outside the stated molecule.

A potential-energy curve gives the same idea a visual form. As two atoms approach, attraction lowers potential energy until a minimum is reached at the equilibrium bond length. Bringing them closer than this creates strong repulsion, while separating them to a very large distance requires energy equal to the bond dissociation energy. A deeper minimum represents a stronger bond; its horizontal position gives the bond length.

A coordinate bond forms when a lone-pair donor supplies both electrons to an electron-deficient acceptor. For example, ammonia donates a lone pair to ${\mathrm{H}}^{+}$ to form $\mathrm{N}{\mathrm{H}}_4^{+}$. Once formed, the coordinate bond is equivalent to the other N–H covalent bonds in the ion.

Coordinate bonding also explains dimeric aluminium chloride, in which chlorine lone pairs bridge between electron-deficient aluminium centres. In an arrow representation, draw the arrow from the atom with the lone pair to the atom or ion that accepts it.

Simple covalent molecules usually have low melting and boiling temperatures because only intermolecular attractions are overcome on melting or boiling; the strong covalent bonds within each molecule remain intact. They usually do not conduct electricity because they contain no mobile ions or delocalised electrons.

Giant covalent structures are different: covalent bonds extend throughout the solid, so they can have very high melting temperatures. Always identify whether a question concerns bonds within particles or forces between particles; confusing the two produces weak property explanations.

A Lewis or dot-and-cross diagram is a bookkeeping model: count the valence electrons, place shared pairs between bonded atoms, then place remaining electrons as lone pairs. The drawing is only acceptable when every atom has the required electron count—or when an expanded octet or odd-electron species is explicitly allowed by the structure being discussed.

Orbital language adds a three-dimensional explanation. A sigma bond comes from direct overlap along the internuclear axis and permits rotation if no other bond prevents it. A pi bond comes from sideways overlap of parallel p orbitals; it is weaker than a sigma bond and restricts rotation because the p orbitals must remain aligned. A double bond therefore contains one sigma and one pi bond, while a triple bond contains one sigma and two pi bonds.

Melting behaviour, conductivity and solubility provide evidence about structure. No one observation proves a model on its own, so compare a pattern of properties against the predictions for molecular, ionic, metallic and giant covalent materials.

Electron bookkeeping before drawing prevents missing lone pairs, incorrect charges and impossible valences.