16.09.02 · inorgchem / lanthanides-actinides

Actinide chemistry and the nuclear fuel cycle

shipped3 tiersLean: none

Anchor (Master): Cotton & Wilkinson 1988 Advanced Inorganic Chemistry (Wiley) Ch. 20; Katz, Seaborg & Morss 1986 The Chemistry of the Actinide Elements (Chapman & Hall); Nash & Madic 2006 Actinide Separation Science and Technology; IAEA Nuclear Fuel Cycle reports

Intuition Beginner

The actinides are the bottom row of the f-block: actinium (Ac, Z = 89) through lawrencium (Lr, Z = 103). Their 5f electrons reach outward, unlike the deeply buried 4f electrons of the lanthanides above them 16.09.01. That single difference reshapes their chemistry. Because the 5f orbitals poke out far enough to overlap neighbouring atoms, the early actinides — uranium, neptunium, plutonium — form bonds their lanthanide cousins cannot, and they flip between several oxidation states instead of sitting stuck at $+ 3$ .

The signature of this chemistry is the actinyl ion. When uranium reaches its $+ 6$ state it becomes the uranyl cation $UO_{2}^{2 +}$ , a linear O=U=O unit with two short, strong bonds. Plutonium and neptunium do the same. These actinyl ions then gather a belt of water, nitrate, or carbonate ligands around their middle, giving the pentagonal- and hexagonal-bipyramidal shapes seen nowhere else in the periodic table.

Why does this matter? Because uranium and plutonium power reactors and arm weapons, and every step of their use — mining, enrichment, irradiation, reprocessing, waste — turns on a redox switch between oxidation states. Separating uranium from plutonium in spent fuel, the PUREX process, is a controlled flip of plutonium between $+ 4$ and $+ 3$ . Storing the leftover minor actinides for millennia is a problem in actinide coordination chemistry. The actinides are where inorganic chemistry meets nuclear engineering.

Visual Beginner

The figure has three panels. The left panel shows the linear actinyl unit O=An=O with two covalent An=O bonds along one axis and a ring of five or six equatorial ligands (water, nitrate) in the perpendicular plane, forming a bipyramid. The middle panel shows the oxidation-state ladder for the early actinides (U, Np, Pu reaching $+ 6$ and $+ 7$ ), with the PUREX switch marked: U(VI) stays put while Pu is cycled Pu(IV) to Pu(III). The right panel traces the fuel cycle as a loop — mining and milling, conversion to $UF_{6}$ , enrichment, fuel fabrication, irradiation in the reactor, cooling, and the back end: PUREX reprocessing or direct disposal, then vitrified waste or spent fuel to a geological repository.

Worked example Beginner

The enrichment separation factor. Natural uranium is two isotopes: uranium-235 (0.72 percent, the fissile fuel) and uranium-238 (99.27 percent, merely fertile). To fuel a typical power reactor you must raise the $^{235} U$ fraction to about 3 to 5 percent. Enrichment does this by converting uranium to the gas uranium hexafluoride, $UF_{6}$ , and exploiting the tiny mass difference between the two isotopologues.

The heavy isotopologue, $^{238} UF_{6}$ , has molar mass $238 + 6 \times 19 = 352$ grams per mole. The light one, $^{235} UF_{6}$ , has $235 + 6 \times 19 = 349$ . Lighter molecules move a little faster, so the ideal single-stage separation factor is the square root of their mass ratio: $α = 352/349 \approx 1.0043$ .

So each enrichment stage lifts the $^{235} U$ fraction by only 0.43 percent. To climb from 0.72 percent to reactor grade, you must chain hundreds to thousands of stages into a cascade. This tiny factor is why enrichment plants are vast industrial complexes — and why the same cascade, run longer, can be pushed to weapons-grade uranium above 90 percent.

Check your understanding Beginner

Formal definition Intermediate+

The actinides are the 15 elements actinium (Ac, $Z = 89$ ) through lawrencium (Lr, $Z = 103$ ), across which the $5 f$ subshell fills from $5 f^{0}$ (Ac) to $5 f^{14}$ (Lr). The valence configuration is formally $5 f^{a} 6 d^{b} 7 s^{2}$ with $a + b$ rising across the series. The defining contrast with the lanthanides, established in 16.09.01, is radial: in the early actinides (Ac–Am) the $5 f$ orbitals extend to radii comparable to the $6 d$ and $7 s /7 p$ valence shell, so they overlap ligand orbitals and participate in bonding; from curium onward the $5 f$ orbitals contract and the chemistry converges toward the lanthanide $+ 3$ norm ^{[CottonWilkinson1988]}.

The actinyl cation is the linear dioxo unit $[O = An = O]^{n +}$ (An $=$ U, Np, Pu, Am) in which the actinide sits at the $+ 5$ or $+ 6$ oxidation state. Its bonding is a superposition of $σ$ and $π$ donation from the two oxo O $2 p$ orbitals into empty actinide $5 f$ and $6 d$ acceptor orbitals of matching symmetry, producing short An=O distances (about $1.75$ – $1.80$ Å for U=O in $UO_{2}^{2 +}$ ) consistent with formal double-bond character. Equatorial coordination by five or six donors (water, nitrate, carbonate) perpendicular to the O=An=O axis gives pentagonal- or hexagonal-bipyramidal geometry ^{[GreenwoodEarnshaw1997]}.

The accessible oxidation states trace the radial story:

Ac^{3 +}, Th^{4 +}, Pa^{5 +}, U^{3 + \dots 6 +}, Np^{3 + \dots 7 +}, Pu^{3 + \dots 7 +}, Am^{2 + \dots 6 +}, Cm^{3 +}, \dots, Lr^{3 +} .

The $5 f$ -driven redox ladder of the early actinides (U, Np, Pu, Am) is what makes their aqueous coordination chemistry rich and what the nuclear fuel cycle exploits.

The ideal single-stage isotope separation factor for gaseous diffusion of an isotopic pair with molar masses $M_{H} > M_{L}$ is

α = \frac{M _{H}}{M _{L}},

which for the $^{235} UF_{6}$ / $^{238} UF_{6}$ pair gives $α = 352/349 \approx 1.0043$ per stage ^[Wilson1996]. The smallness of $α$ is the engineering reason enrichment is a multi-stage cascade.

Key mechanism [Intermediate+] {#key-mechanism}

The mechanism that organises actinide separations is the PUREX redox switch ^[NashMadic]. Spent nuclear fuel is dissolved in nitric acid, sending uranium into solution as the uranyl ion $UO_{2}^{2 +}$ (U(VI)) and plutonium as $Pu^{4 +}$ (Pu(IV)). A kerosene solution of the neutral extractant tributyl phosphate, TBP, $(C_{4} H_{9} O)_{3} P = O$ , is contacted with the aqueous phase. Both U(VI) and Pu(IV) partition into the organic phase as neutral, lipophilic nitrato complexes:

UO_{2}^{2 +} + 2 NO_{3}^{-} + 2 TBP ⇌ UO_{2} (NO_{3})_{2} \cdot 2 TBP,

Pu^{4 +} + 4 NO_{3}^{-} + 2 TBP ⇌ Pu (NO_{3})_{4} \cdot 2 TBP .

The nitrate anions neutralise the cation charge; the two phosphoryl oxygen donors of TBP complete the equatorial coordination sphere on the actinide; the resulting assembly is neutral and hydrocarbon-soluble, so it moves into the organic phase. The trivalent fission-product and lanthanide cations, lacking a charge-matched neutral nitrato complex of comparable stability, stay behind in the aqueous nitric acid.

Separation of uranium from plutonium is then a single reduction step. Pu(IV) is selectively reduced to Pu(III) — by U(IV), hydroxylamine, or ferrous sulfamate — while U(VI) is left untouched:

Pu^{4 +} + e^{-} ⟶ Pu^{3 +}, E^{\circ} (Pu^{4 +} / Pu^{3 +}) \approx + 1.01 V .

The trivalent plutonium would require three nitrates and three TBP donors to assemble a neutral species, a coordination it does not adopt stably; $Pu^{3 +}$ therefore refuses to extract and drops back into the aqueous phase, leaving U(VI) behind in the organic phase. Uranium is subsequently stripped by dilute acid. The whole separation is a controlled flip of plutonium's oxidation state — $+ 4$ co-extracts with uranium, $+ 3$ strips away — and it succeeds because the actinide redox ladder makes Pu(IV) and Pu(III) both stable, kinetically accessible, and chemically distinct.

Bridge. This redox-switch mechanism builds toward every advanced actinide separation downstream — the minor-actinide partitioning that targets Np, Am, and Cm, the grouped actinide extraction (GANEX) envisaged for a fully closed fuel cycle — and appears again in the coordination chemistry of redox-active metals 16.04.01, where oxidation state dictates extractability exactly as it does here. This is exactly the foundational reason PUREX has run essentially unchanged since 1954: the $+ 3$ / $+ 4$ / $+ 6$ actinide redox ladder, set by $5 f$ participation 16.09.01, gives a clean chemical handle that no purely physical separation can match. The central insight is that separation is achieved not by a physical barrier but by a reversible electron-transfer reaction; the bridge is that the same redox ladder powers enrichment chemistry, fuel fabrication, and waste conditioning alike.

Exercises Intermediate+

Exercise 2 (easy, numeric).

A PUREX loaded organic phase contains the uranyl nitrato complex $UO_{2} (NO_{3})_{2} \cdot 2 TBP$ . Taking oxygen as $- 2$ and the nitrato nitrogens as appropriate, verify that uranium is in the $+ 6$ oxidation state, and state how many donor atoms coordinate uranium in this complex (count the two oxo O atoms plus the nitrate O atoms bound to U).

Hint

Use the actinyl charge: $2 O$ at $- 2$ plus U gives the $+ 2$ charge on $UO_{2}^{2 +}$ . Then count how many nitrate oxygens sit in the equatorial belt.

Answer

Charge balance on $UO_{2}^{2 +}$ : $U + 2 (- 2) = + 2$ , so $U = + 6$ (U(VI), the uranyl ion). In the extracted complex the two bidentate nitrates donate four oxygen atoms into the equatorial belt, and the two TBP phosphoryl oxygens donate two more, giving a hexagonal-bipyramidal environment: 2 axial oxo O $+$ 4 equatorial nitrate O $+$ 2 TBP O $=$ 8 donor atoms around uranium. This eight-coordinate, equatorially saturated structure is exactly why the neutral complex is lipophilic and extracts into kerosene.

Exercise 3 (medium, short-answer).

Plutonium is unusual in that $Pu^{3 +}$ , $Pu^{4 +}$ , $PuO_{2}^{+}$ , and $PuO_{2}^{2 +}$ can all coexist in appreciable amounts in the same acidic solution. Explain why, using the relevant standard reduction potentials:

E^{\circ} (Pu^{4 +} / Pu^{3 +}) \approx + 1.01 V, E^{\circ} (PuO_{2}^{+} / Pu^{4 +}) \approx + 1.04 V, E^{\circ} (PuO_{2}^{2 +} / PuO_{2}^{+}) \approx + 0.94 V .

Hint

Look at the gaps between consecutive potentials. When the potential for reducing Pu(V) to Pu(IV) exceeds the potential for reducing Pu(IV) to Pu(III), what reaction becomes thermodynamically favourable?

Answer

The three consecutive potentials are nearly degenerate (within about $0.1$ V of each other). Because $E^{\circ} (PuO_{2}^{+} / Pu^{4 +}) \approx + 1.04 V > E^{\circ} (Pu^{4 +} / Pu^{3 +}) \approx + 1.01 V$ , the disproportionation $3 Pu^{4 +} + 2 H_{2} O ⇌ Pu^{3 +} + 2 PuO_{2}^{+} + 4 H^{+}$ has a small but positive driving force, and the further equilibrium $2 PuO_{2}^{+} ⇌ Pu^{4 +} + PuO_{2}^{2 +}$ is also accessible. The near-equality of the potentials means no single Pu oxidation state is strongly preferred, so Pu(III), Pu(IV), Pu(V), and Pu(VI) all sit within thermodynamic reach of one another and coexist. This is unique to plutonium and is a direct consequence of the $5 f$ -controlled redox ladder; it is also why controlling Pu's oxidation state in PUREX (forcing it to $+ 4$ to extract, then to $+ 3$ to strip) requires careful redox buffering.

Exercise 4 (medium, numeric/symbolic).

An ideal enrichment cascade raises the $^{235} U$ assay from the natural feed $x_{F} = 0.0072$ to a reactor-grade product $x_{P} = 0.035$ (3.5 percent). Using the ideal stage relation

\frac{x _{P} / ( 1 - x _{P} )}{x _{F} / ( 1 - x _{F} )} = α^{N}, α \approx 1.0043,

estimate the number of stages $N$ . (Use $ln (1.0043) \approx 0.00428$ .)

Hint

Compute the two abundance ratios $x / (1 - x)$ , divide them, take the natural logarithm, and divide by $ln α$ .

Answer

Abundance ratio for product: $0.035/0.965 = 0.03627$ . Abundance ratio for feed: $0.0072/0.9928 = 0.007252$ . Ratio of ratios: $0.03627/0.007252 = 5.005$ . Then $N = ln (5.005) / ln (1.0043) = 1.6104/0.00428 \approx 376$ stages. An ideal symmetric cascade therefore needs on the order of $350$ – $400$ stages to reach LEU; a real plant, with tails stripping and non-ideal mixing, needs somewhat more. Pushing to weapons-grade ( $x_{P} = 0.90$ ) would require roughly $1660$ ideal stages — five times as many — which is the engineering signature of a weapons programme.

Exercise 5 (medium, short-answer).

After PUREX has removed uranium and plutonium, the remaining acidic raffinate still contains the trivalent minor actinides ( $Np^{3 +}$ , $Am^{3 +}$ , $Cm^{3 +}$ ) mixed with the trivalent lanthanide fission products ( $Ln^{3 +}$ ). Explain why separating $An^{3 +}$ from $Ln^{3 +}$ is so difficult, and what subtle chemical difference modern partitioning reagents (such as the N-donor ligands BTP and BTBP used in SANEX) exploit.

Hint

Both families are $3 +$ and hard. Consider the difference in the radial extent and covalency of the $4 f$ versus $5 f$ orbitals, and how that affects a slightly softer nitrogen donor versus a hard oxygen donor.

Answer

$An^{3 +}$ and $Ln^{3 +}$ are both trivalent, both hard (oxophilic), and very close in ionic radius (the early $Ln^{3 +}$ and the $An^{3 +}$ overlap almost exactly). Classical oxygen-donor extractants therefore cannot tell them apart. The one distinguishing feature is that the $5 f$ orbitals of the actinides extend slightly further than the buried $4 f$ of the lanthanides and so participate weakly but measurably in covalent bonding. Soft nitrogen-donor ligands such as the bistriazinyl-pyridines (BTP) and bistriazinyl-bipyridines (BTBP) bind the marginally more covalent, slightly softer $An^{3 +}$ a little more strongly than the strictly ionic $Ln^{3 +}$ , giving separation factors of a few units per stage. This is the SANEX strategy: it directly exploits the $5 f$ -versus- $4 f$ bonding difference that defines the whole actinide family.

Exercise 6 (hard, short-answer).

A 1 tonne (heavy metal) batch of uranium fuel irradiated to 33 GWd/t is discharged from the reactor. Estimate the approximate composition of the spent fuel by mass, and identify which fraction dominates the radiotoxicity (i) at 10 years after discharge and (ii) at 1000 years after discharge. Hence state what partitioning-and-transmutation achieves.

Hint

Typical PWR spent fuel is roughly 95 percent residual uranium, about 1 percent plutonium, about 0.1 percent minor actinides, and 3–4 percent fission products. Compare the half-lives of $^{137} Cs$ and $^{90} Sr$ (~30 y) against $^{239} Pu$ , $^{241} Am$ , $^{243} Am$ (centuries to millennia).

Answer

Typical composition: about 95–96 percent unfissioned uranium (now depleted to ~0.8 percent $^{235} U$ , with some $^{236} U$ ), about 0.9–1.0 percent plutonium, about 0.1 percent minor actinides (Np, Am, Cm), and about 3–4 percent fission products. (i) At 10 years, the fission products $^{137} Cs$ ( $t_{1/2} = 30.1$ y) and $^{90} Sr$ ( $t_{1/2} = 28.8$ y) dominate both the heat output and the radiotoxicity. (ii) By 1000 years, those 30-year fission products have decayed through more than 30 half-lives, and the transuranium actinides ( $^{239} Pu$ , $t_{1/2} \approx 2.4 \times 1 0^{4}$ y; $^{241} Am$ , $432$ y; $^{243} Am$ , $7.4 \times 1 0^{3}$ y) dominate the radiotoxicity, which then persists for on the order of $1 0^{5}$ years.

Partitioning-and-transmutation separates the minor actinides from the waste and fissions them (in a fast reactor or accelerator-driven system) into shorter-lived fission products. This can, in principle, compress the time-window over which the waste remains more radiotoxic than the original uranium ore from roughly $1 0^{5}$ years down to a few centuries — turning a geological-timescale problem into an engineered-storage one.

Exercise 7 (hard, short-answer).

Gas centrifuge enrichment has largely replaced gaseous diffusion. Explain, in terms of the separation mechanism, why a centrifuge achieves a far higher single-stage separation factor than the ideal $α \approx 1.0043$ of diffusion, and why the total energy consumption of a centrifuge cascade is one to two orders of magnitude lower.

Hint

Diffusion separates by molecular speed alone (a mass effect scaling as $M_{H} / M_{L}$ ). A centrifuge imposes an external field. Think about what sets the equilibrium radial concentration gradient of a gas in a cylinder spinning at very high angular velocity.

Answer

In gaseous diffusion, separation comes only from the difference in molecular thermal speeds, giving the Graham-law factor $α = M_{H} / M_{L} \approx 1.0043$ , which is close to unity because the relative mass difference of the two $UF_{6}$ isotopologues is only 0.86 percent.

In a gas centrifuge, the $UF_{6}$ gas spins at very high peripheral speed ( $400$ – $700 m s^{- 1}$ ) inside a rotor. The thermodynamic equilibrium in the strong centrifugal field is a Boltzmann distribution in the effective potential, so the radial concentration gradient depends on the molecular mass exponentially rather than as a square root: the heavier $^{238} UF_{6}$ concentrates at the wall and the lighter $^{235} UF_{6}$ near the central axis, with a single-stage separation factor of order $1.1$ – $1.5$ depending on the peripheral speed. Because each stage separates far more, far fewer stages are needed, and because no work is wasted forcing gas through a porous barrier under pressure, the specific energy consumption (separative work per unit product) is one to two orders of magnitude below that of a diffusion cascade. This is why diffusion plants (Oak Ridge K-25, Pierrelatte, Capenhurst) have been retired and centrifuge cascades (e.g. the Zippe-type machines) now dominate enrichment worldwide.

Lean formalization Intermediate+

This unit has lean_status: none and carries no Lean module. Actinide chemistry and the nuclear fuel cycle are bodies of curated measurement — standard redox potentials, actinyl bond lengths, fission cross-sections, isotopic half-lives, spent-fuel inventories — together with semi-empirical models (Graham-law separation factors, PUREX distribution ratios, radiotoxicity decay sums) that consume those measurements as inputs. The one fragment that is genuinely a derivation (the ideal enrichment separation factor $α = M_{H} / M_{L}$ from kinetic theory, formalised in the Full proof set below) is a one-line consequence of the equipartition theorem and does not require a dedicated Mathlib module. A genuinely useful formal layer would be a typed record of (actinide, oxidation state, actinyl geometry, equatorial coordination, An=O distance, E°) plus verified checkers for charge-balance and isotope-decay bookkeeping; such a layer is not present in Mathlib and lies outside the scope of this unit. See the unit metadata Mathlib gap analysis for the full statement.

Advanced results Master

The electronic structure of the actinyl bond

The actinyl cation is the structural and electronic signature of high-oxidation-state actinide chemistry. A molecular-orbital analysis of linear $D_{\infty h}$ $UO_{2}^{2 +}$ shows that the ten valence orbitals of the two oxo oxygens (three $σ$ -type, four $π$ -type, and the non-bonding combinations) interact with the actinide $5 f$ , $6 d$ , $7 s$ , and $7 p$ manifolds to produce a set of $σ_{g, u}$ and $π_{g, u}$ bonding and antibonding combinations, with the strongly bonding $σ_{u}$ and $π_{u}$ levels dominated by oxygen and the corresponding antibonding partners carrying actinide $5 f$ and $6 d$ character ^{[CottonWilkinson1988]}.

The crucial chemical point is that the $5 f_{σ}$ and $5 f_{π}$ orbitals of uranium overlap the oxygen $2 p$ donors well enough to contribute appreciable covalent character to the U=O bond, whereas the deeply buried lanthanide $4 f$ orbitals cannot overlap at all. This $5 f$ participation is the foundational reason the actinyl bond exists, and it is reinforced by relativistic effects: in elements as heavy as uranium, the $1 s$ and inner-shell electrons reach a substantial fraction of the speed of light, the relativistic mass increase contracts the $s$ and $p$ orbitals (which have density at the nucleus), the increased shielding destabilises the $5 f$ and $6 d$ orbitals, and the resulting $5 f$ / $6 d$ energy match with oxygen $2 p$ strengthens the actinyl bond and shortens it beyond what a non-relativistic calculation would predict. Quantitative actinide chemistry therefore requires relativistic quantum chemistry (Dirac–Hartree–Fock and two-component DFT); non-relativistic methods get the bond lengths and vibrational frequencies wrong by several percent.

Isotope separation cascades and separative work

The tiny single-stage factor $α \approx 1.0043$ for gaseous diffusion of $UF_{6}$ makes enrichment a cascade problem. The ideal-symmetric cascade obeys the abundance-ratio relation $(x_{P} / (1 - x_{P})) / (x_{F} / (1 - x_{F})) = α^{N}$ , giving roughly $376$ ideal stages to reach 3.5 percent LEU and roughly $1660$ ideal stages to reach 90 percent weapons-grade uranium. Real plants must also strip the tails to a waste assay $x_{W}$ (typically $0.2$ – $0.3$ percent), which increases the stage count and introduces the concept of separative work ^[Wilson1996].

Separative work, measured in separative work units (SWU, $1 SWU = 1 kg SWU$ ), is the thermodynamic "amount of separation" performed, defined through the value function $V (x) = (1 - 2 x) ln ((1 - x) / x)$ . The SWU balance for a plant producing mass $P$ of product at assay $x_{P}$ from feed $F$ at assay $x_{F}$ with tails $W$ at assay $x_{W}$ is $S = P V (x_{P}) + W V (x_{W}) - F V (x_{F})$ . This formulation, due to Cohen and Dirac, lets enrichment economics be computed independently of the particular separation technology. A centrifuge cascade achieves a given SWU at one to two orders of magnitude lower electrical energy than a diffusion cascade, because the centrifuge's separation factor scales exponentially with peripheral speed (through the Boltzmann factor in the centrifugal potential) rather than as a mass square root.

Spent-fuel composition and the radiotoxicity time-scale

The composition of light-water-reactor spent fuel at discharge is, per tonne of initial heavy metal at about $33 GWd/t$ burn-up: roughly $95$ – $96$ percent residual uranium (depleted to $\sim 0.8$ percent $^{235} U$ , with $^{236} U$ and trace $^{232} U$ ), about $0.9$ – $1.0$ percent plutonium (the fissile $^{239} Pu$ plus $^{238, 240, 241, 242} Pu$ ), about $0.1$ percent minor actinides ( $^{237} Np$ , $^{241, 243} Am$ , $^{242, 244} Cm$ ), and $3$ – $4$ percent fission products (the $30$ -year heat sources $^{137} Cs$ and $^{90} Sr$ , plus long-lived $^{99} Tc$ , $^{129} I$ , $^{135} Cs$ , and the lanthanides) ^[IAEANFC].

The radiotoxicity (ingestion dose potential per unit mass) has two regimes. For the first $300$ – $500$ years, the fission products dominate, principally $^{137} Cs$ and $^{90} Sr$ . Beyond that window, having decayed through many half-lives, the fission products fall below the transuranics, and the actinides $^{239} Pu$ , $^{240} Pu$ , $^{241} Am$ , $^{243} Am$ , and the curium isotopes dominate out to roughly $1 0^{5}$ years. It takes that long for the spent fuel's radiotoxicity to fall back to the level of the natural uranium ore from which it was mined. This time-scale is the engineering basis of geological disposal.

Partitioning, transmutation, and waste forms

The back end of the fuel cycle forks into two strategies. The once-through (open) cycle treats the spent fuel assembly itself as the waste form: the $UO_{2}$ ceramic matrix, sealed inside a corrosion-resistant canister (the Swedish–Finnish KBS-3 concept uses a cast-iron inner container with a 5 cm copper outer shell, embedded in a bentonite-clay buffer in crystalline bedrock), is sent directly to a deep geological repository. The closed cycle reprocesses the fuel by PUREX to recover uranium and plutonium for recycle (the plutonium as mixed-oxide, MOX, fuel), leaving a high-level liquid waste of fission products and minor actinides that is calcined and vitrified into a borosilicate glass log sealed in a stainless-steel canister ^[NashMadic].

Partitioning and transmutation (P&T) is the proposed extension of the closed cycle: partition the minor actinides out of the PUREX raffinate and transmute them by neutron irradiation. The partitioning step is chemically hard because $An^{3 +}$ and $Ln^{3 +}$ are isoelectronic in charge and near-identical in radius (the SANEX/DIAMEX/GANEX family of N-donor extractants exploits the residual $5 f$ covalency to separate them, as in Exercise 5). The transmutation step fissions the minor actinides — for example $^{241} Am$ captures a neutron and ultimately fissions — converting long-lived actinide isotopes into shorter-lived fission products. Done comprehensively, P&T can compress the time for the waste's radiotoxicity to return to uranium-ore levels from $\sim 1 0^{5}$ years to a few centuries, at the cost of building fast-spectrum reactors or accelerator-driven systems and operating a much more demanding chemical separation plant. Advanced ceramic waste forms such as SYNROC (a titanate-based assemblage designed to immobilise actinides and rare-earths individually in separate mineral phases) target the residues of such a P&T cycle.

Synthesis. The actinides resolve into a single scheme driven by the radial extent of the $5 f$ shell: $5 f$ participation enables variable oxidation states and the covalent actinyl bond, the oxidation-state ladder enables PUREX redox separations, the isotope mass difference enables (and constrains) enrichment, and the residual actinide inventory sets the back-end waste problem. This is exactly the unification that Katz, Seaborg and Morss build across the actinide series; the foundational reason the actinides are chemically rich where the lanthanides are uniform is the radial reach of the $5 f$ orbital. Putting these together generalises actinide chemistry into the same coordination-and-redox engine that runs the rest of the periodic table 16.04.01, and the bridge is that the $5 f$ -driven redox ladder builds toward PUREX, enrichment, and partitioning-and-transmutation, while appears again in the ligand-field and solid-state chemistry of 16.03.x and 16.07.x, and the periodic-trend machinery of 16.01.01.

Full proof set Master

Proposition (Ideal single-stage isotope separation factor). Let $M_{H} > M_{L}$ be the molar masses of the heavy and light isotopologues of a gas in equilibrium at temperature $T$ . The ratio of the molecular fluxes through a small aperture (the ratio of effusion rates) equals $α = M_{H} / M_{L}$ . For $^{238} UF_{6}$ versus $^{235} UF_{6}$ , $α = 352/349 \approx 1.00429$ .

Proof. From the equipartition theorem, the mean kinetic energy of each molecular species in thermal equilibrium is $\frac{1}{2} m ⟨ v^{2} ⟩ = \frac{3}{2} k_{B} T$ , so the mean squared speed is $⟨ v^{2} ⟩ = 3 k_{B} T / m$ and the mean speed $\overset{v}{ˉ} \propto 1/ m \propto 1/ M$ . The number of molecules crossing a unit area of aperture per unit time — the effusion flux $J$ — equals the number density times the mean speed, $J = n \overset{v}{ˉ} /4$ (the factor $1/4$ is the standard kinetic-theory angular average). Taking the ratio of light to heavy fluxes at equal number densities gives

α = \frac{J _{L}}{J _{H}} = \frac{v ˉ _{L}}{v ˉ _{H}} = \frac{M _{H}}{M _{L}} .

Substituting $M_{H} = 352 g/mol$ and $M_{L} = 349 g/mol$ ,

α = \frac{352}{349} = 1.008596 = 1.00429.

Each ideal stage of gaseous diffusion therefore enriches the light isotopologue by a factor of only $1.00429$ , or about $0.43$ percent. This is Graham's law of effusion, and it is the upper bound on what a single diffusion stage can achieve; no amount of clever barrier design raises $α$ above $M_{H} / M_{L}$ for an ideal gas. $■$

Corollary (Cascade stage count). For an ideal symmetric enrichment cascade operating from feed assay $x_{F}$ to product assay $x_{P}$ , the number of stages $N$ satisfies

\frac{x _{P} / ( 1 - x _{P} )}{x _{F} / ( 1 - x _{F} )} = α^{N}, N = \frac{ln [ \frac{x _{P} ( 1 - x _{F} )}{x _{F} ( 1 - x _{P} )} ]}{ln α} .

For natural uranium ( $x_{F} = 0.0072$ ) enriched to reactor-grade LEU ( $x_{P} = 0.035$ ), $N \approx 376$ ideal stages; enriched to weapons-grade uranium ( $x_{P} = 0.90$ ), $N \approx 1664$ ideal stages.

Proof. In an ideal symmetric cascade with no mixing losses, the abundance ratio $r = x / (1 - x)$ is multiplied by $α$ at each stage, so after $N$ stages $r_{P} = α^{N} r_{F}$ . Solving for $N$ gives $N = ln (r_{P} / r_{F}) / ln α$ . For LEU: $r_{P} = 0.035/0.965 = 0.03627$ , $r_{F} = 0.0072/0.9928 = 0.007252$ , $r_{P} / r_{F} = 5.005$ , $N = ln (5.005) / ln (1.00429) = 1.6104/0.004279 = 376.3$ . For weapons-grade: $r_{P} = 0.90/0.10 = 9$ , $r_{P} / r_{F} = 9/0.007252 = 1241$ , $N = ln (1241) /0.004279 = 7.1248/0.004279 = 1665$ . Real cascades with tails stripping and non-ideal mixing require somewhat more stages, but the ideal count fixes the order of magnitude and shows why weapons-grade enrichment demands roughly five times as many stages as reactor-grade. $■$

Connections Master

Lanthanides and actinides — the f-block 16.09.01 supplies the periodic-trend framing this unit builds on: the $4 f$ -versus- $5 f$ radial contrast, the lanthanide contraction, and the overview of actinide oxidation states. This unit descends from that overview into the applied actinide chemistry — actinyl bonding, PUREX, enrichment, the fuel cycle — that the periodic-trend treatment could only gesture at, and the $5 f$ participation invoked throughout is the same radial-extendibility argument made quantitative.
Coordination chemistry 16.04.01 supplies the coordination-number, geometry, and chelate-effect vocabulary applied here to actinyl equatorial coordination (pentagonal/hexagonal bipyramids, eight-coordinate U(VI) nitrato complexes) and to the N-donor ligands of minor-actinide partitioning. The redox-switch separations of this unit are the actinide-specific instance of the oxidation-state-controls-chemistry principle that organises all of 16.04.x.
Crystal-field and ligand-field splitting 16.03.02 is the framework whose actinide extension (with $5 f$ and $6 d$ manifolds replacing the pure $3 d$ set) describes the actinyl electronic structure of this unit's Advanced results. The linear $D_{\infty h}$ actinyl MO diagram is the analogue, one row deeper and with relativistic corrections, of the $O_{h}$ crystal-field splitting of 16.03.02.
Solid-state chemistry 16.07.01 and 16.07.03 governs the waste-form end of the fuel cycle: the $UO_{2}$ fluorite lattice of spent fuel, the borosilicate glass and SYNROC ceramic matrices that immobilise high-level waste, and the defect and radiation-damage chemistry that controls actinide retention in a geological repository over $1 0^{5}$ years.
Periodic trends quantified 16.01.01 provides the effective-nuclear-charge and ionic-radius data that underlie the actinide/lanthanide radius comparison exploited by minor-actinide partitioning, and the relativistic-orbital-contraction argument that strengthens the actinyl bond. The actinide redox potentials used throughout are read directly off the periodic-trend machinery of 16.01.x.

Historical & philosophical context Master

The actinide chemical industry was born under the pressure of the Second World War. The discovery of plutonium — synthesised by bombardment of uranium with deuterons by Glenn Seaborg, Edwin McMillan, Joseph Kennedy, and Arthur Wahl in 1940–41, with its chemistry worked out in extraordinary haste ^{[Seaborg1946]} — turned the actinides from a curiosity into a strategic material within five years. The Hanford B reactor produced the first gramme-scale quantities of plutonium in 1944 using a bismuth-phosphate coprecipitation process (carry-through of $Pu^{4 +}$ with $BiPO_{4}$ , leaving $Pu^{3 +}$ in solution by selective reduction), the direct ancestor of the redox-switch idea that PUREX would systematise. PUREX itself — tributyl phosphate in a hydrocarbon diluent extracting U(VI) and Pu(IV) nitrato complexes — was demonstrated at the Knolls Atomic Power Laboratory and brought to industrial scale at the Savannah River and Hanford plants in the 1950s ^{[Flanary1956]}; it has been the workhorse reprocessing chemistry ever since, essentially unchanged in principle for seventy years.

The enrichment side of the fuel cycle has its own arc. Gaseous diffusion, scaled up at the Oak Ridge K-25 plant during the Manhattan Project, was the dominant enrichment technology from 1945 through the 1980s, despite requiring colossal electrical input because of the $M_{H} / M_{L}$ limit on each stage. The Zippe-type gas centrifuge, developed in the Soviet Union by a team including captured German engineer Gernot Zippe in the 1950s and then refined in the West, exploited the exponential mass dependence of the Boltzmann distribution in a centrifugal field to achieve far higher separation per stage and far lower energy use; by the turn of the twenty-first century diffusion plants had been retired worldwide and centrifuge cascades dominated.

The back end of the cycle — waste — is the part still unresolved. The United States, after President Carter's 1977 moratorium on commercial reprocessing, adopted a once-through policy and committed to the Yucca Mountain geological repository, a programme that has yet to accept any commercial spent fuel. France, the United Kingdom, Russia, Japan, and India reprocess commercially by PUREX and vitrify the residue; Finland's Onkalo repository, under construction in Precambrian bedrock at Olkiluoto since the 1990s, is the first deep-geological-disposal facility to begin licensing for spent fuel. The partitioning-and-transmutation programme — advanced fast reactors and accelerator-driven systems aimed at fissioning the minor actinides — is the modern research front, targeting the reduction of the $1 0^{5}$ -year radiotoxicity tail that neither direct disposal nor conventional PUREX vitrification addresses ^[IAEANFC]. For the standard technical survey of the whole cycle from ore to wastes, see Wilson's The Nuclear Fuel Cycle ^[Wilson1996]; for the actinide chemistry at the depth this unit reaches, see Katz, Seaborg and Morss's The Chemistry of the Actinide Elements ^{[CottonWilkinson1988]} and Nash and Madic on actinide separations ^[NashMadic].

Bibliography Master

Cotton, F. A. & Wilkinson, G. (1988). Advanced Inorganic Chemistry, 6th ed. Wiley-Interscience, New York. Chapter 20 — the actinides: oxidation states, actinyl $[AnO_{2}]^{n +}$ bonding, aqueous redox potentials, descriptive and coordination chemistry.
Greenwood, N. N. & Earnshaw, A. (1997). Chemistry of the Elements, 2nd ed. Butterworth-Heinemann, Oxford. Chapter 31 — actinide chemistry, the nuclear fuel cycle, uranium enrichment, and actinyl ions.
Katz, J. J., Seaborg, G. T. & Morss, L. R. (eds.) (1986). The Chemistry of the Actinide Elements, 2nd ed., 2 vols. Chapman & Hall, London. The canonical reference for $5 f$ chemistry, actinyl electronic structure, plutonium disproportionation, and solvent extraction.
Nash, K. L. & Madic, C. (2006). Actinide Separation Science and Technology. In Morss, L. R., Edelstein, N. M., Fuger, J. & Katz, J. J. (eds.), The Chemistry of the Actinide and Transactinide Elements, 3rd ed., Chapter 24. Springer, Dordrecht. PUREX, minor-actinide/lanthanide partitioning (DIAMEX, SANEX, GANEX), and N-donor ligand selectivity.
Choppin, G. R., Liljenzin, J.-O. & Rydberg, J. (2013). Radiochemistry and Nuclear Chemistry, 4th ed. Elsevier, Oxford. Actinide aqueous chemistry, redox potentials, PUREX chemistry, and isotope separation.
Wilson, P. D. (1996). The Nuclear Fuel Cycle: From Ore to Wastes. Oxford University Press, Oxford. Mining, conversion, enrichment, fuel fabrication, irradiation, reprocessing, and waste conditioning.
International Atomic Energy Agency (various years). Nuclear Fuel Cycle Information System and IAEA TecDoc series on spent-fuel composition, radiotoxicity inventories, partitioning-and-transmutation, and geological disposal. IAEA, Vienna.
Seaborg, G. T., Wahl, A. C. & Kennedy, J. W. (1946). Radioactive element 94 from deuterons on uranium. Physical Reviews 69, 367. The plutonium discovery and its early chemistry.
Flanary, J. R. (1956). Solvent extraction system for uranium and plutonium recovery from irradiated fuel. Proceedings of the International Conference on the Peaceful Uses of Atomic Energy 9, 479–485. The industrial-scale PUREX process.
Pepper, D. E. (1981). Uranium enrichment technology and management. Chemical Engineering Progress 77, 57–66. Gaseous diffusion, the gas centrifuge, and the separative-work framework.

Prerequisites

16.09.01
16.04.01

Tier anchors

beginner: Housecroft & Sharpe 2018 Inorganic Chemistry (Pearson) Ch. 24 (the actinide actinyl ions, the nuclear fuel cycle as the applied face of actinide redox)
intermediate: Miessler, Fischer & Tarr 2014 Inorganic Chemistry (Pearson) Ch. 23 (5f radial extension, actinide oxidation-state ladder, actinyl bonding, PUREX)
master: Cotton & Wilkinson 1988 Advanced Inorganic Chemistry (Wiley) Ch. 20; Katz, Seaborg & Morss 1986 The Chemistry of the Actinide Elements (Chapman & Hall); Nash & Madic 2006 Actinide Separation Science and Technology; IAEA Nuclear Fuel Cycle reports

References

Cotton, F. A. & Wilkinson, G. — Advanced Inorganic Chemistry, 6th ed. (Wiley, 1988) · Ch. 20 — actinide oxidation states, actinyl [AnO2]n+ bonding, aqueous redox potentials, separations
Greenwood, N. N. & Earnshaw, A. — Chemistry of the Elements, 2nd ed. (Butterworth, 1997) · Ch. 31 — actinide descriptive chemistry, uranium enrichment, PUREX, the nuclear fuel cycle
Katz, J. J., Seaborg, G. T. & Morss, L. R. — The Chemistry of the Actinide Elements, 2nd ed. (Chapman & Hall, 1986) · Vol. 1–2: 5f chemistry, actinyl electronic structure, plutonium disproportionation, solvent extraction
Nash, K. L. & Madic, C. — Actinide Separation Science and Technology; in The Chemistry of the Actinide and Transactinide Elements, 3rd ed. (Springer, 2006) · Ch. 24 — PUREX, minor-actinide/lanthanide partitioning (DIAMEX, SANEX, GANEX), N-donor ligand selectivity
International Atomic Energy Agency — Nuclear Fuel Cycle Information System, IAEA TecDoc series (Vienna) · Spent-fuel composition, radiotoxicity inventories, partitioning-and-transmutation, geological disposal
Wilson, P. D. — The Nuclear Fuel Cycle: From Ore to Wastes (Oxford University Press, 1996) · Mining, conversion, enrichment, fuel fabrication, irradiation, reprocessing, waste conditioning
Choppin, G. R., Liljenzin, J.-O. & Rydberg, J. — Radiochemistry and Nuclear Chemistry, 4th ed. (Elsevier, 2013) · Actinide aqueous chemistry, redox potentials, PUREX chemistry, isotope separation

Estimated time

beginner: 20 min
intermediate: 55 min
master: 110 min