Phosphorus Halides and Hypervalency
Drop a lump of white phosphorus into a stream of dry chlorine gas. The reaction is almost violent. A pale yellow-green fog forms within seconds. You get either a fuming liquid (PCl₃), or, with excess chlorine, a straw-colored solid. That solid hisses and fumes the moment it meets damp air (PCl₅). One of these molecules is a modest pyramid. The other has five bonds radiating from a single phosphorus atom. This defies the simple octet rule that works so well for carbon and nitrogen. That single fact — five bonds, ten "electrons" around one center — turned phosphorus pentahalides into the textbook battleground for the idea of hypervalency. This question occupied chemists from Gilbert Lewis through Linus Pauling to the modern computational era. This matters because PX₃ and PX₅ are not museum curiosities. PCl₃ is manufactured on a megaton scale. It is the entry point to almost the entire organophosphorus industry. This includes pesticides, flame retardants, and the PCl₃ → POCl₃ → phosphate ester routes. These routes end up in everything from motor oil additives to nerve-agent precursors (and their antidotes). PF₅ and PCl₅ are the reagents of choice for converting alcohols and carboxylic acids to halides. Their extreme moisture sensitivity is a direct, mechanistic consequence of the same electronic structure. That structure is what makes them "hypervalent" in the first place. Understanding why PCl₅ is trigonal bipyramidal in the gas phase requires certain tools. The same tools explain why it is ionic ([PCl₄]⁺[PCl₆]⁻) in the solid. The same is true for why it hydrolyzes stepwise through POCl₃ to phosphoric acid. This article builds exactly those tools: VSEPR geometry, a correct (non-3d-orbital) molecular orbital picture of the P–X bonding, and an appreciation of how halogen electronegativity and size govern both structure and reactivity. The deeper theoretical story is a genuine correction of twentieth-century chemical folklore. Generations of students were taught that PCl₅ "uses" phosphorus 3d orbitals. This was said to accommodate ten valence electrons. This idea seemed to explain hypervalency, but quantitative calculations have since dismantled it. The modern, defensible picture replaces d-orbital hybridization with three-center four-electron (3c–4e) bonding. It also involves negative hyperconjugation, and ionic/resonance contributions. These concepts are shared with noble-gas compounds, sulfur hexafluoride, and the interhalogens. This is not a semantic quibble. The model you choose predicts different bond lengths, different vibrational spectra, and different reactivities. Only one of the models survives contact with the data. This article works through the structures of PX₃ and PX₅ from first principles. It builds the 3c–4e bonding picture explicitly with orbital diagrams. It derives the trigonal bipyramidal preference from electron-pair repulsion. It also uses second-order perturbation theory for this derivation. It then uses that electronic picture to explain — not just describe — the stepwise hydrolysis chemistry. This chemistry converts these halides into oxoacids and esters. This sets up the wider chemistry of Nitrogen and Phosphorus Oxoacids.
- PX₃ (X = F, Cl, Br, I) is pyramidal (C₃ᵥ), lone pair on P, consistent with AX₃E VSEPR; basicity and Lewis-base behavior follow directly from that lone pair.
- PX₅ is trigonal bipyramidal (D₃ₕ) in the gas/liquid phase for PF₅ and PCl₃, with distinct axial and equatorial bond lengths (axial longer, weaker).
- "Hypervalency" in PCl₅ is NOT explained by phosphorus 3d-orbital hybridization; the defensible model uses 3-center–4-electron bonding across the axial P–Cl–Cl framework plus ionic character.
- Solid PCl₅ is ionic, [PCl₄]⁺[PCl₆]⁻ (tetrahedral cation, octahedral anion) — a direct consequence of the weaker axial bonds and Berry pseudorotation flexibility.
- Both PX₃ and PX₅ hydrolyze readily; PX₃ gives H₃PO₃ (phosphorous acid) + HX, while PX₅ gives POX₃ then H₃PO₄ + HX, via nucleophilic attack (associative, SN2(P)-like) at phosphorus.
- Key idea: apparent 10-electron "hypervalency" at phosphorus is real in terms of bond count but is built from ordinary s/p orbitals redistributed via 3-center bonding and ionic resonance, not from d-orbital participation.
Phosphorus never truly exceeds an s+p octet of shared electron pairs in the orbital sense; apparent five- and six-coordination is achieved through long, weak, electron-rich 3-center–4-electron bonds (and ionic lattices), not through 3d-orbital hybridization.
Where this lives today
PCl₃ remains one of the highest-tonnage phosphorus compounds manufactured worldwide, produced directly from white phosphorus and chlorine gas and consumed almost entirely as a synthetic intermediate. Its largest downstream use is oxidation to POCl₃ (phosphoryl chloride), which then reacts with alcohols to give trialkyl phosphates — plasticizers and flame retardants used in everything from PVC to circuit boards — and with amines or further substitution chemistry to build the P=O and P–N backbones found in organophosphate pesticides and, notoriously, in the same chemical family as nerve agents, which is why PCl₃ manufacture and trade are monitored under chemical weapons non-proliferation frameworks even though the overwhelming majority of PCl₃ output goes into benign agricultural and industrial products. PCl₅ and PF₅ are workhorse reagents in synthetic organic and inorganic chemistry for converting carboxylic acids to acyl chlorides and alcohols to alkyl halides, exploiting exactly the electrophilicity at phosphorus and the facility of associative substitution described above. PF₅ additionally serves as a strong Lewis acid catalyst and as a fluorinating/activating agent in specialty organic synthesis, its stability as a purely molecular (non-ionizing) pentahalide making it more tractable to handle than PCl₅ in some contexts. The bonding models developed here extend directly into modern materials and biological chemistry. The 3-center–4-electron picture used for the axial P–Cl–Cl unit of PCl�5 is the same conceptual tool used to describe the hypervalent phosphorus centers proposed as intermediates in phosphoryl transfer enzymes (kinases and phosphatases), where a nucleophile, the phosphorus center, and a leaving group briefly form an analogous three-center arrangement during in-line associative or dissociative transition states — directly relevant to understanding ATP hydrolysis and DNA/RNA phosphodiester backbone chemistry. The same hypervalent bonding logic (VSEPR shape plus 3c–4e electronic detail rather than d-orbital hybridization) is now the standard treatment for SF₆, ClF₃, I₃⁻, XeF₂ and other classic "expanded octet" species across p-Block Chemistry, making phosphorus pentahalides the canonical teaching case for a bonding principle with a genuinely broad footprint across the periodic table.
Ask the tutor: How does the 3-center–4-electron bonding picture change when one axial position in a PCl₅ derivative is replaced by a different, less electronegative group — does the whole molecule reorganize, or just the axial bond order?
The physical picture
Start with PX₃. Phosphorus has the configuration [Ne]3s²3p³, five valence electrons. Bonding to three halogens uses three of those electrons in P–X bonds, leaving one lone pair. VSEPR treats this as an AX₃E system: four electron domains around phosphorus, arranged tetrahedrally, with three occupied by bonding pairs and one by the lone pair. Because the lone pair occupies more angular space than a bonding pair (it is held by one nucleus only, so its density is less compressed), the X–P–X angles compress from the ideal tetrahedral 109.5° to values in the 96–100° range depending on halogen. PF₃ has an F–P–F angle of about 97.8°; PCl₃ about 100.1°; PBr₃ about 101.0°; PI₃ about 102.0°. The lone pair sits at the apex of a trigonal pyramid, is stereochemically active, and is the reason PX₃ compounds are good σ-donor (and, for PF₃, π-acceptor) ligands to transition metals — exactly the lone pair that makes PPh₃ and PF₃ classic ligands in coordination chemistry. Now go to PX₅. Only PF₅ and PCl₅ (and, more exotically, PBr₅ under special conditions) exist as genuine five-coordinate molecular species; PI₅ is not known as a stable neutral pentaiodide because iodine is too large and too poor a σ-donor to stabilize five P–I bonds — steric and electronic factors both cap the halogen size that pentacoordination will tolerate. In the gas and liquid phases, PF₅ and PCl₃-derived PCl₅ adopt the trigonal bipyramidal (TBP) geometry, point group D₃ₕ: three equatorial substituents at 120° to each other in a plane containing phosphorus, and two axial substituents perpendicular to that plane at 90° to every equatorial position and 180° to each other. This is the unique five-coordinate geometry that minimizes electron-pair repulsion for five identical ligands (as opposed to a square pyramid, which is a higher-energy transition state connecting two TBP arrangements). Crucially, axial and equatorial P–X bonds are not equivalent. In PCl₅, gas-phase electron diffraction gives axial P–Cl = 2.14 Å and equatorial P–Cl = 2.02 Å — axial bonds are about 6% longer and correspondingly weaker (lower bond order). In PF₅, axial P–F = 1.58 Å versus equatorial P–F = 1.53 Å, the same qualitative pattern. This asymmetry is the single most important experimental clue for how bonding actually works in these molecules, and it rules out any model that treats all five P–X bonds as equivalent two-center-two-electron single bonds built from five equivalent hybrid orbitals.
The asymmetry is explained by the 3-center–4-electron (3c–4e) bonding model applied to the axial X–P–X unit. Treat the phosphorus 3pz orbital (oriented along the axial direction) and the two axial halogen p orbitals pointing at phosphorus as a three-orbital basis. Combining these three atomic orbitals gives three molecular orbitals: a fully bonding combination (all same phase), a nonbonding combination (mostly localized on the two halogens, with a node at phosphorus), and a fully antibonding combination. Four electrons — two from the P–X σ bonds that would naively be drawn, recast in this basis — fill the bonding and nonbonding MOs, leaving the antibonding MO empty. The result is a formal bond order of 0.5 per P–X(axial) linkage: two axial bonds share the strength of a single ordinary bond pair, exactly matching the observed lengthening and weakening of the axial bonds relative to the equatorial ones, which remain normal 2c–2e σ bonds built from sp² hybrids in the equatorial plane. This is the same bonding motif used to describe XeF₂, I₃⁻, and the axial fluorines of SF₄ — it is a general feature of hypervalent p-block chemistry, not a phosphorus peculiarity.
The core ideas
The key quantitative relationships governing these molecules are geometric, energetic, and electronic. First, the bond order in the 3c–4e axial system:
where: and are electrons in bonding and antibonding MOs of the 3-orbital, 4-electron axial fragment; the nonbonding MO contributes no bond order since it has a node at phosphorus. Reading: with 2 electrons in the bonding MO, 2 in the nonbonding MO, 0 in the antibonding MO, and two P–X linkages sharing that bonding pair, each axial bond gets bond order . Second, the relationship between electronegativity of the halogen and the stability/existence of PX₅, captured qualitatively by the apicophilicity and σ*-donor balance:
where: is the Pauling electronegativity of the halogen and its covalent radius; larger stabilizes the electron-rich 3c–4e axial bonding (better able to hold partial negative charge in the nonbonding MO) while larger destabilizes the crowded five-coordinate arrangement. Reading: fluorine (highest χ, smallest r) gives the most stable PF₅; iodine (lowest χ among the halides that form PX₃, largest r) cannot support a stable PI₅ — both electronic and steric terms work against it simultaneously. Third, the energetics of the solid-state disproportionation of PCl₅ into an ionic lattice:
is favorable in the solid because lattice energy released by forming a well-matched ionic solid ([PCl₄]⁺ tetrahedral, [PCl₆]⁻ octahedral, both symmetric, both closed hypervalent-free or cleanly hypervalent structures) exceeds the cost of charge separation — but in the gas phase, where no lattice energy exists, molecular D₃ₕ PCl₅ is the stable form. This is a clean illustration of how phase changes the operative bonding description.
Goal: show from electron-pair repulsion (VSEPR) that D₃ₕ TBP is the electrostatically preferred arrangement of five equivalent electron pairs around a central atom, and quantify why axial positions differ from equatorial.
Step 1 — count the ligand–ligand repulsion terms. For five points on a sphere around a central atom, repulsion energy summed pairwise (Coulomb's law-like falloff with angle) depends only on the angles between all pairs of positions. In an ideal TBP, angles are: 3 equatorial–equatorial pairs at 120°, 6 axial–equatorial pairs at 90°, and 1 axial–axial pair at 180°. In a square pyramid, four basal ligands sit at 90° to each other and to the apex ligand roughly at 90–105°, giving many more 90° close contacts.
Step 2 — weight repulsions by angle. Because repulsion falls off steeply with angle (roughly as some inverse power of the angle or its cosine in simple electrostatic models), 90° contacts dominate total repulsion energy; 120° contacts are noticeably less repulsive, and 180° contacts are negligible. The TBP has 6 contacts at 90°; the square pyramid, with five ligands crowded into a base-plus-apex arrangement, generates 8 contacts at 90° or less (four basal–basal at 90°, four apical–basal at closer than 90° if the apex is pulled toward the base to relieve other strain). Fewer 90° contacts means lower total repulsion.
Step 3 — locate substituent electronegativity effects. The two geometrically distinct sites in a TBP (axial vs equatorial) experience different numbers of 90° neighbors: an equatorial position has only two 90° neighbors (the two axial ligands), while an axial position has three 90° neighbors (the three equatorial ligands). Lower-electronegativity substituents, or lone pairs (which are "fatter" and repel more), preferentially occupy the equatorial site where they face fewer 90° close contacts — this is the origin of the general rule that lone pairs and less electronegative/bulkier groups go equatorial, while highly electronegative or small substituents tolerate the more crowded axial site.
Step 4 — connect to bond lengths. Because axial ligands are simultaneously squeezed by three 90° equatorial neighbors, the axial bonds are inherently longer/weaker even for a homoleptic PX₅ — consistent with the 3c–4e picture, which independently predicts a bond order of 0.5 for axial bonds. The geometric (VSEPR) argument and the orbital (3c–4e) argument converge on the same qualitative prediction, which is why both are taught together: VSEPR gets the shape right for the wrong (or incomplete) electronic reason, and MO theory supplies the correct electronic mechanism behind that shape.
Go deeper · Derivation B · 2 minWhy 3d orbitals are not needed — the hyperconjugation/ionic alternative to d-orbital hybridization
Goal: show quantitatively why the historical "sp³d² hybridization" picture for PCl₅ fails, and construct the defensible alternative.
Step 1 — energetic mismatch. Phosphorus 3d orbitals lie roughly 10–15 eV above the 3s/3p valence orbitals in the free atom (spectroscopically, P 3d excited states are far above the ground-state manifold). Effective hybridization requires comparable orbital energies so the resulting hybrid is a genuine, low-energy mixture; mixing an orbital 10+ eV higher in energy contributes negligibly to a bonding combination, because the energy denominator in perturbation theory,
is small whenever is large, suppressing the 3d coefficient regardless of the numerator.
Step 2 — population analysis contradicts d-orbital occupation. High-level ab initio calculations (natural bond orbital / NBO analyses) on PF₅ and PCl₅ consistently find d-orbital populations on phosphorus of only a few hundredths of an electron — far too small to be called a "hybridized" bonding orbital, and consistent with d functions acting merely as polarization functions that improve the basis set's flexibility rather than as genuinely occupied valence orbitals.
Step 3 — build the alternative from s and p only. Phosphorus has four valence orbitals (3s, 3pₓ, 3p_y, 3p_z) available for bonding, exactly enough for four two-electron bonds if used conventionally — but PCl₅ needs five. The resolution: use sp² hybrids (from 3s + 3pₓ + 3p_y) for three ordinary 2c–2e bonds to the three equatorial chlorines, and use the remaining pure 3p_z orbital, together with the two axial chlorine p-orbitals, to build the 3-center-4-electron system described above. No fifth phosphorus orbital, and certainly no d orbital, is required — the "extra" bonding capacity comes from spreading 4 electrons over 3 centers rather than manufacturing a genuinely new hybrid orbital.
Step 4 — connect to negative hyperconjugation and ionic resonance. The nonbonding MO in the 3c–4e picture is heavily weighted on the two axial chlorines: physically, this describes significant negative charge buildup on the axial halogens, equivalent to writing resonance structures with partial Cl⁻ character and a P⁺ center donating into a Cl–P σ* orbital of the opposite axial bond (negative hyperconjugation, analogous to the anomeric effect). Taken to its ionic extreme, this is exactly the driving force behind the [PCl₄]⁺[PCl₆]⁻ ionic lattice observed in the solid: the axial bonding was already "ionic-flavored" in the molecular form, and crystal packing simply completes the charge separation when lattice energy makes it worthwhile.
Worked examples
Example 1 — Foundational: predicting and rationalizing the shape and bond angle of PBr₃. Phosphorus tribromide: draw the Lewis structure. Phosphorus contributes 5 valence electrons, each bromine contributes 7, giving a total of valence electrons for the neutral molecule. Three P–Br single bonds use 6 electrons (3 bonding pairs); each bromine needs 3 lone pairs to complete its octet, using electrons; the remaining electrons form one lone pair on phosphorus. Electron domain count at P: 3 bonding pairs + 1 lone pair = 4 domains, AX₃E, so the electron-domain geometry is tetrahedral and the molecular shape is trigonal pyramidal (C₃ᵥ). Because the lone pair is more diffuse and repels more strongly than bonding pairs, the Br–P–Br angle compresses below the ideal 109.5°; experimentally it is 101.0°, consistent with the general trend PF₃ (97.8°) < PCl₃ (100.1°) < PBr₃ (101.0°) < PI₃ (102.0°) — as the halogen becomes less electronegative, the P–X bonding pair is pulled less toward the halogen, spends more time near phosphorus, and therefore repels the other bonding pairs more, pushing the angle closer to (though still under) the ideal tetrahedral value. Limiting case: if the lone pair were completely absent (hypothetically, an AX₄ system like PBr₄⁺), the angle would relax fully to 109.5°, which is indeed observed for the isoelectronic, lone-pair-free tetrahedral cation [PCl₄]⁺.
Example 2 — Applied: quantifying the axial/equatorial bond order asymmetry from vibrational/structural data in PF₅. Gas-phase electron diffraction gives P–F(axial) = 1.577 Å and P–F(equatorial) = 1.534 Å in PF₅. Using a simple empirical bond-order/bond-length correlation of the Pauling type,
where is taken as a reference single-bond P–F length (use the equatorial bond, bond order 1, as the reference: Å), solve for the axial bond order :
This empirical estimate (~0.87) is not identical to the idealized 3c–4e value of exactly 0.5, because real axial P–F bonds are strengthened somewhat by ionic contraction (fluorine's small size and high electronegativity let it pull real electron density into a more conventional-looking bond, partially offsetting the 3c–4e weakening) and because the simple Pauling equation is a rough empirical fit, not a first-principles bond-order calculator. The qualitative conclusion — axial bond order measurably below 1, matching the 3c–4e prediction of substantial weakening relative to a normal single bond — holds regardless of the exact numerical value extracted from any one empirical formula. Limiting case: if axial and equatorial bonds were truly identical (as in a naive "5 equivalent sp³d² hybrids" picture), would equal and this calculation would return exactly — it does not, which is itself the falsification of the d-hybrid model.
Example 3 — Exam/competition level: full stepwise hydrolysis stoichiometry and mechanism for PCl₅, with mass-balance and mechanistic justification. PCl₅ reacts with water in two distinguishable stages depending on stoichiometry and conditions. With limited water (1 equivalent), the reaction stops at phosphoryl chloride:
Check the mass balance: left side has 1 P, 5 Cl, 2 H, 1 O; right side has 1 P, (3 + 2) = 5 Cl, 2 H, 1 O. Balanced. With excess water, POCl₃ (and any remaining PCl₅) hydrolyzes fully to phosphoric acid:
Mass balance: left side 1 P, 3 Cl, 6 H, 4 O; right side 1 P, 3 Cl, (3+3)=6 H, 4 O. Balanced. Combining the two steps, the overall reaction of PCl₅ with excess water (4 equivalents total) is:
Mass balance: left 1 P, 5 Cl, 8 H, 4 O; right 1 P, 5 Cl, (3+5)=8 H, 4 O. Balanced. Mechanistically, each step is nucleophilic substitution at phosphorus. Water's oxygen lone pair attacks the electrophilic phosphorus center — made electrophilic by the electronegative chlorines withdrawing electron density — most readily at an equatorial-type or axial site depending on the transient geometry, forming a pentacoordinate (or, for the initial PCl₅ attack, hexacoordinate transition state/intermediate) that then expels chloride as HCl after proton transfer. Because phosphorus in PCl₅ is already coordinatively saturated at five, water addition first generates a fleeting six-coordinate dianion-like transition state analogous to the [PCl₆]⁻ ion seen in the solid-state lattice, which collapses by loss of Cl⁻ (immediately protonated to HCl by the departing water-derived proton) to give four-coordinate POCl₃. The second and third hydrolysis steps on POCl₃ proceed by the same associative mechanism at a now four-coordinate (tetrahedral) phosphorus center, passing through five-coordinate trigonal bipyramidal transition states — exactly the geometry this article has already justified as the low-energy shape for five-coordinate phosphorus — with the incoming water entering apically and the best leaving group (Cl⁻) also departing apically, consistent with the general preference for nucleophilic attack and leaving-group departure to occur from axial-like positions in phosphorus substitution chemistry. Limiting case: if only 1 equivalent of water is present per PCl₅ and the reaction is quenched rapidly at low temperature, POCl₃ can be isolated as the kinetic product, exactly as industrially exploited; only with a large water excess or extended reaction time does the second and third hydrolysis proceed to completion, giving H₃PO₄.
Limiting cases, scope & common traps
| Situation | What happens | Why |
|---|---|---|
| PX₃, small/electronegative X (F) | Most pyramidal, smallest angle (97.8°) | Bonding pairs pulled toward F, less P-centered repulsion |
| PX₃, large/less electronegative X (I) | Least pyramidal, largest angle (102°) | Bonding pairs closer to P, more repulsion, angle opens |
| PX₅ with X = F, Cl | Stable molecular D₃ₕ TBP (gas/liquid) | High χ stabilizes 3c–4e axial system; small r tolerates crowding |
| PX₅ with X = I | Not isolable as neutral PI₅ | Low χ destabilizes electron-rich axial bonding; large r sterically forbidden |
| PCl₅ in the solid state | Ionic: [PCl₄]⁺[PCl₆]⁻ | Lattice energy of a well-matched ionic solid exceeds cost of charge separation |
| PCl₅ in nonpolar solvents / gas phase | Molecular D₃ₕ | No lattice stabilization available; ionic form not favored |
| PF₅ in any common phase | Always molecular D₃ₕ | Smaller ionic radii mismatch, weaker tendency to ionize; F⁻ too small/hard to favor forming a comparable ionic lattice under normal conditions |
| Axial vs equatorial substitution preference | Electronegative/small groups go axial; lone pairs, bulky/less electronegative groups go equatorial | Fewer 90° repulsive contacts in equatorial sites; apicophilicity rules |
| Explaining hypervalency with 3d orbitals | Historically taught, now considered incorrect as a primary bonding description | 3d orbitals lie too high in energy; NBO populations on d orbitals are negligible |
The chemist behind it
The theoretical understanding of hypervalency did not emerge from a single laboratory but from a long argument stretching across the twentieth century, with three figures standing out as central protagonists.
Origins · 1 minGilbert N. Lewis and how this was discovered
Lewis introduced the shared electron-pair bond and the octet rule in 1916, giving chemistry its first rigorous electronic bonding language. His own structures for hypervalent species like PCl₅ already required going beyond a strict octet, and Lewis himself was cautious about over-generalizing the octet rule, recognizing that heavier main-group elements could accommodate more than eight electrons around a central atom — an empirical accommodation later mis-rationalized by others as necessarily requiring d-orbital participation.
Pauling's hybridization framework was an extraordinarily successful piece of chemical theory for ordinary octet-obeying molecules, and his extension of it to sp³d and sp³d² hybrids for PCl₅ and SF₆ was a reasonable, testable hypothesis given the tools available in the 1930s. It took forty more years of structural refinement, and then another twenty years of computational quantum chemistry, to show that the hypothesis, while geometrically suggestive, was electronically wrong in its literal orbital-mixing claim — one of the cleaner examples in inorganic chemistry of a widely taught model being superseded not by a change in the observed facts (the TBP shape was correct all along) but by a deeper and more accurate accounting of why the shape occurs.
Why does PCl₅ hydrolyze so violently compared to, say, CCl₄?
Phosphorus in PCl₅ is coordinatively unsaturated in a reactive sense: even though it already has five bonds, it can readily expand to a six-coordinate transition state because trigonal bipyramidal and octahedral geometries are close in energy and interconvertible via Berry pseudorotation-like pathways. Water's oxygen lone pair attacks the strongly electrophilic phosphorus (electron density is pulled away by five electronegative chlorines) with a low activation barrier, and the reaction is also strongly exothermic because P=O and P–OH bonds are much stronger than P–Cl bonds. Carbon in CCl₄, by contrast, has no accessible low-energy pathway to five-coordination (no low-lying d-like acceptor orbitals, and a five-coordinate carbon transition state is prohibitively high in energy), so CCl₄ hydrolyzes far more slowly, essentially not at all under ordinary conditions.
Is it ever correct to draw PCl₅ with an expanded octet using d orbitals, even as a simplification?
As a bookkeeping device for counting electrons and predicting shape via VSEPR, drawing ten electrons around phosphorus is a useful shorthand and does not itself claim anything about orbital character — VSEPR's predictive success does not depend on the mechanism by which the "extra" bonding is achieved. The problem arises only if the ten-electron Lewis structure is used to assert that phosphorus literally hybridizes 3d orbitals into the bonding scheme; that specific mechanistic claim is not supported by modern calculations and should not be taught as the electronic explanation, even though the resulting geometric prediction (TBP) remains correct.
Why is PF₃ toxic and notorious for binding hemoglobin, similar to carbon monoxide?
PF₃ has a phosphorus lone pair (σ-donor) combined with low-lying P–F σ* orbitals that make it a reasonably good π-acceptor, closely paralleling the bonding of CO to metal centers, including the iron in hemoglobin. This donor-acceptor synergy lets PF₃ bind iron(II) in heme very strongly, competitively displacing O₂ and blocking oxygen transport, exactly analogous to CO poisoning, though PF₃ is far less common as an environmental hazard than CO because it is not a combustion byproduct.
Does the 3-center-4-electron model apply to PF₅ exactly the same way as to PCl₅?
Qualitatively yes: the axial fluorines occupy a 3c–4e system analogous to the chlorines in PCl₅, and the same axial-longer/equatorial-shorter bond length pattern is observed. Quantitatively, fluorine's small size and very high electronegativity let it accommodate the partial negative charge in the nonbonding MO more comfortably, and empirical bond-order estimates for axial P–F bonds come out somewhat higher than the idealized 0.5 (closer to 0.85–0.9 by simple Pauling-type bond-length correlations), reflecting real chemical differences in how "ideal" the 3c–4e picture is for a small, hard halogen versus a larger, softer one.
Why does PCl₅ hydrolysis stop cleanly at POCl₃ with limited water, rather than giving a statistical mixture of partially hydrolyzed products?
The first hydrolysis step is fast and strongly exothermic because it forms a very strong P=O bond, effectively locking in the tetrahedral POCl₃ product and outcompeting any redistribution among partially substituted species. With only 1 equivalent of water available, essentially all of it is consumed converting PCl₅ to POCl₃ plus HCl before any further hydrolysis of the P–Cl bonds in POCl₃ (which requires additional water and is comparatively slower due to POCl₃'s lower electrophilicity than PCl₅) can compete significantly, especially under controlled, cool conditions typical of industrial POCl₃ production.
How does VSEPR correctly predict PX₅ geometry without needing to know anything about 3c–4e bonding?
VSEPR is a purely geometric/electrostatic model: it counts electron domains (bonding pairs and lone pairs) around the central atom and finds the arrangement that minimizes pairwise repulsion, without specifying which atomic or hybrid orbitals house those electron pairs. For five bonding domains and no lone pairs on phosphorus, VSEPR correctly predicts the D₃ₕ trigonal bipyramid purely from repulsion-minimization geometry. The 3-center-4-electron model is a separate, complementary layer that explains why the underlying orbitals redistribute electron density asymmetrically between axial and equatorial positions — VSEPR gets the shape right without needing this deeper electronic detail, which is why it remains pedagogically valuable even after the d-orbital explanation for the bonding is abandoned.
Why does phosphorus form both PX₃ and PX₅ while nitrogen essentially only forms NX₃, never a stable NX₅
Nitrogen has no accessible low-lying d-like or extended-valence pathway analogous to phosphorus's 3-center bonding option, and more fundamentally nitrogen is simply too small: a genuine five-coordinate nitrogen center would require impossibly short N–X bonds and severe steric crowding around the tiny nitrogen atom, with no orbital of appropriate symmetry and energy available to build even a 3c–4e-type system in the second period. Phosphorus's larger covalent radius and the availability of energetically reasonable 3p orbitals oriented for 3-center bonding make PX₅ chemistry possible in a way that has no true second-period analogue — nitrogen pentahalides simply do not exist as stable, isolable species under ordinary conditions.