Skip to content

Radio-analytical Techniques

CSIR NET chemistry: detectors, counting statistics, isotope dilution, radiometric titration, radioimmunoassay, tracers, Szilard–Chalmers and dating.

  • 18 min read
  • Updated
On this page · 18 sections
  1. Prerequisites
  2. The Hook: Weighing What Cannot Be Weighed
  3. How Radiation Interacts With Matter
  4. Radiation Detectors
  5. Counting Statistics
  6. Isotope Dilution Analysis (IDA)
  7. Radiometric Titrations
  8. Radioimmunoassay (RIA)
  9. Tracers in Chemistry
  10. Radiometric Dating
  11. Radiation Units in Brief
  12. Memory Tricks
  13. Don’t Confuse
  14. Common Mistakes
  15. Practice
  16. Quick Revision Table
  17. Key Takeaways
  18. What’s Next

Prerequisites

The Hook: Weighing What Cannot Be Weighed

How Radiation Interacts With Matter

Detectors work because radiation ionises or excites the material it passes through.

RadiationInteractionRangeShielding
αdense ionisation along a short straight track (high linear energy transfer)a few cm of air; stopped by papertrivial externally; very hazardous inside the body
β⁻, β⁺ionisation along a tortuous path; bremsstrahlung for energetic β in heavy materials; β⁺ annihilates to two 511 keV photonsup to metres of air, millimetres of tissue or plasticlow-Z materials (plastic, perspex) first
γ, X-raysphotoelectric effect (low energy, high Z), Compton scattering (intermediate), pair production (above 1.022 MeV)exponential attenuation, no fixed rangehigh-Z, dense materials (Pb, concrete)
Neutronsnuclear reactions and elastic collisions, no direct ionisationlonghydrogenous moderator plus absorber (B, Cd)

For γ spectroscopy the photoelectric effect is what matters: it deposits the whole photon energy in the detector and gives the photopeak used to identify a nuclide. Compton scattering gives the continuous background below it.

Radiation Detectors

DetectorPrincipleBest forEnergy resolution
Ionisation chambercollects primary ion pairs, no gas multiplicationhigh activities, dose calibrators—
Proportional countergas multiplication, pulse ∝ primary ionisationα/β discrimination, low-energy X-raysmoderate
Geiger–Müller (GM) counterfull avalanche: every event gives the same large pulsesimple counting of β and γnone (cannot measure energy); long dead time
NaI(Tl) scintillatorlight pulse ∝ energy, read by a photomultiplierγ counting with high efficiencypoor to moderate
Liquid scintillation counting (LSC)sample dissolved in a scintillator cocktail; no window to crosslow-energy β emitters: ³H, ¹⁴C, ³⁵Smoderate; suffers from quenching
HPGe semiconductorelectron–hole pairs in germanium, cooled with liquid nitrogenγ spectrometry of mixtures, activation analysisbest (sharp photopeaks)
Si(Li), Si surface barrieras above, in siliconX-rays, α spectrometryhigh
Autoradiography / phosphor imaging, film, TLDlatent image or trapped chargelocating labels on gels, tissue; dosimetry—

Why LSC for ³H and ¹⁴C: their β particles are so weak (maximum energies of about 18.6 keV for ³H and 156 keV for ¹⁴C) that they would be absorbed by a counter window or by the sample itself. In LSC the emitter and scintillator are mixed at the molecular level. Coloured or chemically interfering samples reduce light output (quenching), which is corrected with internal standards or quench curves.

Counting Statistics

Radioactive decay is random: the number of counts N recorded in a fixed time follows a Poisson distribution (close to Gaussian for large N), with

$$\boxed{\sigma_N = \sqrt{N}} \qquad \text{relative standard deviation} = \frac{1}{\sqrt{N}}$$
Counts NσRelative σ
1001010%
2500502%
10 0001001%

So the precision depends on the number of counts, not on the count rate or the time separately. To halve the relative error, collect four times as many counts.

Count rates. For N counts in time t, R = N/t and \( \sigma_R = \sqrt{N}/t = \sqrt{R/t} \).

Net rate after background. Errors add in quadrature:

$$R_{\text{net}} = R_g - R_b, \qquad \sigma_{\text{net}} = \sqrt{\frac{R_g}{t_g} + \frac{R_b}{t_b}}$$

For a fixed total counting time, the most precise net rate comes from dividing time so that \( t_g/t_b = \sqrt{R_g/R_b} \).

Dead time. After each event a detector is blind for a time τ (hundreds of microseconds for a GM tube, much shorter for scintillation and semiconductor systems). For a non-paralysable detector, true rate n and observed rate m are related by

$$\boxed{n = \frac{m}{1 - m\tau}}$$

Efficiency and geometry. Count rate is not activity. Activity (Bq) = count rate / (detection efficiency × fraction of emissions of the measured type). Standards counted in the same geometry remove most of these factors, which is why radio-analytical methods are usually comparative.

Isotope Dilution Analysis (IDA)

The most important quantitative radio-analytical method. Its great advantage: the analyte does not have to be recovered quantitatively, only in a pure (if small) portion whose specific activity can be measured.

Direct isotope dilution

Add a known mass W₁ of labelled analyte with specific activity S₁ to the sample containing an unknown mass Wₓ of the same compound. Mix thoroughly so the label is evenly distributed. Isolate any pure portion and measure its specific activity S₂. Total activity is conserved:

$$S_1 W_1 = S_2 (W_1 + W_x) \quad\Longrightarrow\quad \boxed{W_x = W_1\left(\frac{S_1}{S_2} - 1\right)}$$

Variants

VariantWhat is unknownWhat is addedUse
Direct IDAinactive analytelabelled analyteamino acids, metals, organics in mixtures
Reverse (inverse) IDAamount of a labelled substance in a sampleknown mass of inactive carrier of the same compoundyield of a labelled product, metabolite studies: \( W_x = W_c/(S_x/S_m - 1) \)
Double IDAspecific activity of the label itself also unknowntwo aliquots with different inactive additionswhen the tracer is not calibrated
Substoichiometric IDA—equal, deliberately less-than-stoichiometric amounts of reagent added to sample and standard, so equal masses are isolatedtrace metals; then masses cancel and only activities are compared

In substoichiometric IDA the same amount of analyte is isolated from both the standard and the spiked sample (because the reagent is the limiting factor), so the ratio of measured activities gives the answer without weighing the isolated fractions.

Radiometric Titrations

A titration in which the end point is found from radioactivity rather than an indicator. One reactant is labelled, and a phase separation (precipitation, solvent extraction) moves activity between phases at the equivalence point.

Example: chloride titrated with ¹¹⁰ᵐAg-labelled Ag⁺. Before the end point every added Ag⁺ precipitates as AgCl, so the solution activity stays near zero. After the end point excess labelled Ag⁺ stays in solution and the solution activity rises linearly. Two straight lines cross at the equivalence volume.

Labelled speciesSolution activity before e.p.After e.p.Shape
Titrant (Ag⁺*)≈ 0risesflat, then rising
Analyte (Cl⁻*)falls as analyte precipitates≈ 0 (constant)falling, then flat
Both labelledfallsrisesV-shaped

Radiometric titrations work at low concentrations and in coloured or turbid solutions where indicators fail.

Radioimmunoassay (RIA)

Developed by Rosalyn Yalow and Solomon Berson for insulin; Yalow received the Nobel Prize in Physiology or Medicine for it. RIA is a competitive binding assay:

  1. A fixed, limited amount of antibody.
  2. A fixed amount of radiolabelled antigen (often ¹²⁵I-labelled, a γ/X-ray emitter that is easy to count).
  3. The unlabelled antigen from the sample (or a standard).

Labelled and unlabelled antigen compete for the limited binding sites. The more unlabelled antigen in the sample, the less labelled antigen is bound. After separating bound from free antigen (for example with a second antibody or by precipitation), the bound activity, or the ratio bound/free, is read against a standard curve.

graph LR
    A["Sample antigen
(unlabelled)"] --> C{"Limited
antibody"} B["¹²⁵I-antigen
(fixed amount)"] --> C C --> D["Separate
bound / free"] D --> E["Count bound fraction"] E --> F["More sample antigen
→ lower bound counts"]

RIA reaches picogram to nanogram levels of hormones, drugs and vitamins. Non-radioactive immunoassays (ELISA, chemiluminescence) have since replaced it in many routine uses, but the competitive principle is the same. For the antibody side, see Immunotechniques, Histochemistry and Microscopy.

Tracers in Chemistry

Solubility products

Precipitate a salt from a reagent of known specific activity, saturate a solution with it, and count an aliquot of the clear solution. Counts ÷ specific activity = moles dissolved. Hevesy’s lead experiments were of this kind. A worked version is Practice Q6 below.

Mechanisms and structural equivalence

  • Thiosulfate. Boil sulfite with elemental ³⁵S to make thiosulfate, S³⁵SO₃²⁻. On decomposition (for example with acid, or via Ag₂S₂O₃ → Ag₂S), the label appears almost entirely in the sulfur or sulfide product and not in the sulfate or SO₂. So the two sulfur atoms of thiosulfate are not equivalent and do not exchange.
  • Isotope exchange. Mix two chemical forms of an element, one labelled, and follow how fast the label redistributes. A fast exchange suggests a labile pathway (for example electron transfer between two oxidation states), a slow one an inert system. The fraction exchanged F follows the McKay equation, \( -\ln(1 - F) = R\,t\,(a+b)/(ab) \), where a and b are the concentrations of the two forms and R the exchange rate.
  • Self-diffusion in solids and melts, and surface area measurements by exchange, use the same idea.

Stable isotopes are tracers too (¹⁸O in ester hydrolysis, ²H, ¹³C, ¹⁵N), but they are detected by mass spectrometry or NMR rather than counting.

Hot-atom chemistry: the Szilard–Chalmers process

When a nucleus captures a neutron and emits a γ ray, the atom recoils with an energy far larger than a chemical bond energy, so it is usually knocked out of its molecule. In the classic experiment, ethyl iodide is irradiated with neutrons: ¹²⁷I(n,γ)¹²⁸I. The recoiling ¹²⁸I atoms leave the C–I bond and end up as free iodine or iodide, which is shaken out into water, leaving the bulk inactive iodine behind in the organic phase.

Conditions for a useful Szilard–Chalmers separation:

  1. The recoil energy must exceed the bond energy, so the bond breaks.
  2. The free radioactive atoms must not exchange rapidly with the target molecules (or the gain is lost).
  3. The new chemical form must be easy to separate.

The result is a radionuclide of much higher specific activity than the target material, from an (n,γ) reaction that otherwise cannot give carrier-free products (see Nuclear Reactions).

Carriers and purity

TermMeaning
Isotopic carrierstable isotope of the same element added so that a trace radionuclide behaves as a weighable quantity (precipitates, adsorbs normally)
Hold-back carrierstable form of an interfering element, added so its radioactive isotopes stay behind
Scavengera non-specific precipitate (for example Fe(OH)₃) used to sweep out many trace activities
Radiocolloidcarrier-free trace activity that adsorbs onto dust and glass like a colloid instead of behaving as a solute
Radionuclidic purityfraction of total activity due to the stated nuclide
Radiochemical purityfraction of the nuclide’s activity present in the stated chemical form

Radiometric Dating

For a parent P decaying to a stable daughter D, with no daughter present initially and a closed system:

$$D = P\left(e^{\lambda t} - 1\right) \quad\Longrightarrow\quad \boxed{t = \frac{1}{\lambda}\ln\left(1 + \frac{D}{P}\right)}$$
MethodDecayHalf-lifeRange, use
Radiocarbon¹⁴C → ¹⁴N (β⁻)about 5730 yorganic material, up to about 50 000 y
K–Ar (and ⁴⁰Ar/³⁹Ar)⁴⁰K → ⁴⁰Ar (EC branch)about 1.25 × 10⁹ yvolcanic rocks; argon escapes from melts, resetting the clock
Rb–Sr⁸⁷Rb → ⁸⁷Sr (β⁻)very long (tens of billions of years)rocks, by the isochron method, which removes the need to know initial ⁸⁷Sr
U–Pb²³⁸U → ²⁰⁶Pb; ²³⁵U → ²⁰⁷Pb4.47 × 10⁹ y; 7.0 × 10⁸ yzircons; two independent clocks (concordia)

Radiocarbon. ¹⁴C is made in the upper atmosphere by ¹⁴N(n,p)¹⁴C from cosmic-ray neutrons, oxidised to ¹⁴CO₂, and taken up by plants and the food chain. Living matter keeps a steady specific activity (about 15.3 disintegrations per minute per gram of carbon in the classic calibration). After death uptake stops and the activity decays:

$$t = \frac{t_{1/2}}{\ln 2}\ln\frac{A_0}{A}$$

Real ages are corrected with calibration curves because the atmospheric ¹⁴C level has varied, and because fossil-fuel CO₂ and atmospheric weapons tests disturbed it. Accelerator mass spectrometry (AMS) counts ¹⁴C atoms directly in milligram samples.

Radiation Units in Brief

QuantitySI unitOlder unit
Activitybecquerel (Bq) = 1 s⁻¹curie (Ci) = 3.7 × 10¹⁰ Bq
Absorbed dosegray (Gy) = 1 J kg⁻¹rad = 0.01 Gy
Equivalent / effective dosesievert (Sv), weighted for radiation type and tissuerem = 0.01 Sv
Exposure (X, γ in air)C kg⁻¹roentgen

Memory Tricks

Don’t Confuse

ThisvsThat
Direct IDA: labelled analyte added to inactive sampleReverse IDA: inactive carrier added to labelled sample
GM counter: counts only, no energy informationHPGe: high-resolution γ spectrometry
Count rate (cpm)Activity (Bq): rate corrected for efficiency and emission probability
Radionuclidic purity (which nuclide)Radiochemical purity (which chemical form)
Isotopic carrier (same element)Hold-back carrier (interfering element)
RIA: competitive, labelled antigenSandwich immunoassays: labelled second antibody, signal rises with analyte
Statistical error depends on counts NNot on count rate alone

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Numerical data given in a question are for that question.

Level 1: Foundation

  1. A sample gives 10 000 counts. What is the relative standard deviation?
  2. Which detection method suits a ³H-labelled sample?
  3. Name the SI unit of absorbed dose.
Answers
  1. 1/√10 000 = 1%.
  2. Liquid scintillation counting (the β is too weak to pass a window).
  3. Gray (1 J kg⁻¹).

Level 2: Part B Style

Practice Q1. 10.0 mg of ¹⁴C-glycine (5000 cpm mg⁻¹) is added to a protein hydrolysate. Pure glycine isolated after mixing has 400 cpm mg⁻¹. The glycine content of the hydrolysate is:

(a) 125 mg (b) 115 mg (c) 12.5 mg (d) 0.80 mg

Answer: (b)

Wₓ = 10.0(5000/400 − 1) = 115 mg. (a) is the total after spiking.

Practice Q2. A sample gives 4000 counts in 10 min and the background 500 counts in 10 min. The net count rate with its standard deviation is:

(a) 350 ± 6.7 cpm (b) 350 ± 5.9 cpm (c) 350 ± 8.6 cpm (d) 400 ± 6.3 cpm

Answer: (a)

Net = 400 − 50 = 350 cpm; σ = √(4000 + 500)/10 = 6.7 cpm. (b) subtracts the variances (√3500/10 = 5.9); (c) adds the standard deviations ((63.2 + 22.4)/10).

Practice Q3. A GM counter with a dead time of 100 µs records 2000 counts s⁻¹. The true count rate is:

(a) 2000 s⁻¹ (b) 2200 s⁻¹ (c) 2500 s⁻¹ (d) 1667 s⁻¹

Answer: (c)

n = m/(1 − mτ) = 2000/(1 − 2000 × 10⁻⁴) = 2000/0.8 = 2500 s⁻¹. At high rates dead-time losses become large; 20% here.

Practice Q4. In a radiometric titration of chloride with Ag⁺ labelled with ¹¹⁰ᵐAg, the activity of the supernatant solution plotted against volume of titrant:

(a) falls linearly and then stays constant (b) stays near zero and then rises linearly (c) rises linearly and then falls (d) rises linearly throughout

Answer: (b)

Before the end point all labelled Ag⁺ precipitates as AgCl; after it, excess labelled Ag⁺ stays in solution. (a) is the shape when the analyte, not the titrant, is labelled.

Level 3: Part C Style

Practice Q5. Charcoal from an archaeological hearth shows 3.8 disintegrations per minute per gram of carbon. Take the value for living material as 15.3 dpm g⁻¹ and t½(¹⁴C) = 5730 y. Which statements are correct?

  1. The sample is about 11 500 years old.
  2. The age is a little over two half-lives.
  3. Counting a 1 g sample for 100 min (with negligible background) gives a relative counting error of about 5%, if the counter efficiency is 100%.
  4. ¹⁴C in the atmosphere is formed by ¹²C(n,γ)¹³C followed by a second neutron capture.

(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only
  1. t = (5730/0.693) ln(15.3/3.8) = 8267 × 1.393 = 1.15 × 10⁴ y. Correct.
  2. Two half-lives = 11 460 y; the age is slightly larger. Correct.
  3. Counts = 3.8 × 100 = 380; 1/√380 = 5.1%. Correct.
  4. Wrong: ¹⁴C comes from ¹⁴N(n,p)¹⁴C.

Practice Q6. AgCl is precipitated with Ag⁺ of specific activity 2.0 × 10⁹ cpm per mol Ag, washed and shaken with water to saturation. A 1.00 mL aliquot of the clear saturated solution gives 26.8 cpm (background-corrected, efficiency included in the specific activity). The K_sp of AgCl is closest to:

(a) 1.3 × 10⁻⁵ (b) 1.8 × 10⁻¹⁰ (c) 1.8 × 10⁻¹⁶ (d) 3.6 × 10⁻¹⁰

Answer: (b)

Moles of Ag in 1.00 mL = 26.8/(2.0 × 10⁹) = 1.34 × 10⁻⁸ mol, so s = 1.34 × 10⁻⁵ mol L⁻¹. K_sp = s² = 1.8 × 10⁻¹⁰. (a) is s, not K_sp; (c) forgets to convert mL to L (s = 1.34 × 10⁻⁸ taken as mol L⁻¹).

Practice Q7. Which statements about radio-analytical methods are correct?

  1. In direct isotope dilution, losing half the analyte during purification does not change the result, provided the isolated portion is pure.
  2. In substoichiometric isotope dilution, equal amounts are isolated from sample and standard, so only activities need be compared.
  3. In the Szilard–Chalmers process, rapid exchange between recoil atoms and the parent molecules is required for high specific activity.
  4. A NaI(Tl) detector gives better energy resolution than an HPGe detector.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1 and 3 only

Answer: (a) 1 and 2 only
  1. Correct: IDA uses specific activity, not total recovery.
  2. Correct: the limiting reagent fixes the isolated amount.
  3. Wrong: exchange would carry the label back into the parent compound and destroy the enrichment. It must be slow.
  4. Wrong: HPGe has far better resolution; NaI(Tl) is chosen for efficiency and cost.

Quick Revision Table

IdeaFact
γ interactionsphotoelectric (photopeak), Compton, pair production > 1.022 MeV
DetectorsGM: counts only, long dead time; NaI(Tl): efficient γ; HPGe: best resolution; LSC: ³H, ¹⁴C (quenching)
Statisticsσ = √N; relative 1/√N; σ_R = √N/t; σ_net = √(R_g/t_g + R_b/t_b); t_g/t_b = √(R_g/R_b)
Dead timen = m/(1 − mτ)
Direct IDAWₓ = W₁(S₁/S₂ − 1); no quantitative recovery needed
Reverse IDAinactive carrier added to labelled sample
Substoichiometric IDAequal isolated amounts; compare activities
Radiometric titrationlabelled titrant: flat then rise; labelled analyte: fall then flat
RIAcompetitive; ¹²⁵I antigen; bound label falls as analyte rises; Yalow–Berson
Tracerssolubility (Hevesy), thiosulfate S non-equivalence, exchange (McKay), self-diffusion
Szilard–Chalmersrecoil breaks bond; no fast exchange; separable form; high specific activity
Datingt = (1/λ) ln(1 + D/P); ¹⁴C 5730 y, ≈ 15.3 dpm g⁻¹ living; K–Ar, Rb–Sr isochron, U–Pb
UnitsBq, Ci; Gy (J kg⁻¹); Sv

Key Takeaways

What’s Next

Advertisement
Advertisement