Advanced Analytical Methods

Advanced analysis is not about using instruments merely because they are available. The professional question is:

Which remaining hypothesis cannot be resolved by routine methods, and what additional information can change the conclusion?

In diamond gemology, an “advanced method” may mean better spatial resolution, more sensitive chemical analysis, three-dimensional imaging, measurement of the crystal lattice, or destructive microsampling. These advantages are not interchangeable.

Four axes of a method

Before selecting an instrument, at least four properties must be distinguished:

1. detection limit — how small a signal the method can register;
2. spatial resolution — how small a region it can distinguish;
3. sampling depth or volume — the part of the sample from which the signal originates;
4. destructiveness — whether the sample remains physically unchanged.

A highly sensitive method may have poor spatial resolution. A method with excellent resolution may require sampling. “More advanced” therefore does not automatically mean “better for every problem.”

[VISUAL 72.1: Four axes of an analytical method—detection limit / spatial resolution / sampling volume / destructiveness]

Nondestructive, minimally destructive, and destructive

A nondestructive approach has a strong advantage for a valuable faceted diamond. Micro-CT and certain X-ray and optical methods can provide important information without removing material. Other techniques, such as LA-ICP-MS and some SIMS applications, may leave a microscopic sampling mark.

The scientific value of the information must be weighed against the potential consequences of the intervention. For routine commercial identification, an invasive method is very often unjustified; for research on a rare inclusion or geological provenance, it may be essential.

X-ray radiography and micro-CT

Radiography provides a projection image of differences in X-ray absorption. Micro-CT reconstructs a three-dimensional volume from a series of projections.

For diamonds, micro-CT can assist in:

• 3D documentation of cavities and inclusions;
• determining the spatial relationships of internal structures;
• investigating rare “diamond-in-diamond” and similar samples;
• visualizing certain inclusion clouds or zones when sufficient X-ray contrast exists.

Its considerable value does not mean that micro-CT is a standard natural/laboratory-grown test. It is a method for examining internal structure, and interpretation must be guided by the question.

XRD: phase and crystal structure

X-ray diffraction uses the diffraction response of the crystal lattice. In a gemological context, it is particularly useful when a mineral or crystalline phase must be identified and optical or Raman data do not resolve it with sufficient reliability.

XRD does not automatically provide the geological history of an inclusion. Identifying a phase means determining what the material is; only then do petrogenetic questions follow.

XRF and LA-ICP-MS are not the same chemical analysis

XRF is a largely nondestructive method of elemental analysis, especially useful for elements and concentrations within its own sensitivity range. With small localized phases or very low concentrations, it may be limited by geometry and sampling volume.

LA-ICP-MS uses laser ablation of a microscopic amount of material and can reach very low concentrations of many elements with good spatial control. The cost of this sensitivity is microdestructiveness and the need for rigorous calibration, standards, and matrix corrections.

Trace-element geographic-origin data from colored-stone literature must not be transferred automatically to diamond. The problem of diamond geographic origin has its own evidentiary architecture.

[VISUAL 72.2: XRF versus LA-ICP-MS—what they measure, volume, sensitivity, and destructiveness]

SIMS and isotope microanalysis

Secondary ion mass spectrometry enables analysis of very small zones and isotopic compositions. In diamond research, it can link particular growth zones to their carbon, nitrogen, or other isotopic properties.

This is particularly important when a single crystal preserves multiple growth episodes. A bulk value can average out differences that are geologically decisive.

An isotopic signature, however, is not automatically a geographic passport for a particular polished stone.

Cathodoluminescence

Cathodoluminescence excites emission with an electron beam rather than UV photons. It can reveal very fine growth or defect structures that are not equally visible in standard deep-UV imaging.

In more recent research, CL has usefully resolved some CVD specimens whose DiamondView patterns were not classically diagnostic. This illustrates the role of an advanced method well: it does not replace the entire workflow but resolves a specific remaining ambiguity.

EPR and paramagnetic defects

Electron paramagnetic resonance measures certain defects or centers with unpaired electrons. In diamond science, it can provide extremely specific information about defect structure.

But EPR is not a routine “diamond tester.” Instrumentation, preparation, interpretation, and actual relevance depend on the problem. Its strength lies in specialized research, not in replacing FTIR, PL, or a standard laboratory workflow.

X-ray topography: the crystal lattice as a map

X-ray topography can visualize dislocations, strain-related structures, and other inhomogeneities of the crystal lattice. As early as 2004, GIA/Gems & Gemology described a procedure in which a topographic record of lattice defects could serve as a highly stable individual fingerprint of a faceted diamond, including the possibility of matching it again after recutting or certain treatments.

An important qualification from the same work is that the method has not become a widely used routine “passport” system in the trade. It should therefore be treated here as a powerful specialized identification concept, not a standard retail workflow.

It is not the same as a birefringence image under crossed polarizers. Both methods can be sensitive to structural inhomogeneity, but the physical signal and spatial interpretation are not the same.

From a point to a map

Much of modern analysis is moving from a single spectrum toward mapping:

• PL map;
• Raman map;
• chemical map;
• cathodoluminescence image;
• X-ray map;
• multimodal overlay.

If multiple maps of the same stone are registered in a shared coordinate system, one can determine whether growth zones, defect centers, chemistry, and strain coincide. This correlation is often more informative than the strongest individual peak.

[VISUAL 72.3: Multimodal registration—the same zone on PL, CL, strain, and chemical maps]

AI and machine learning are not new physics

AI can classify images, find patterns, integrate multidimensional data, and estimate membership in known populations. But a model learns from the data it receives.

The key questions are therefore:

• what the training set is like;
• whether the test data are truly independent;
• whether they cover new manufacturers and treatment conditions;
• how the system responds to a sample outside the known distribution;
• whether human review exists;
• whether the model is versioned and revalidated after an update.

Domain shift can reduce performance without any change in the instrument’s physics. A good system must therefore know when it is uncertain.

Quality control is part of the measurement

An advanced method without quality control produces sophisticated but weak evidence. The following may be required:

• calibration and a reference standard;
• a blank and a control;
• background correction;
• a matrix-matched standard;
• repeated measurement;
• documentation of precision and accuracy;
• software and library versions;
• reproducibility across instruments or laboratories.

Detection limit is also not the same as quantification limit: a signal may be detected before it can be quantified reliably.

Research-grade and routine commercial service

A laboratory may have scientific expertise in a method that is not part of every commercial service. Reasons may include:

• analysis time;
• cost;
• destructiveness;
• an insufficiently validated routine workflow;
• the need for a rare expert;
• lack of need for typical samples.

Therefore, “the laboratory has the instrument” is not the same as “that instrument is used for every report.”

When escalation is justified

Advanced analysis has the greatest value when:

1. routine findings contradict one another;
2. the consequence of an incorrect conclusion is substantial;
3. the object is scientifically or historically important;
4. a new material or unusual treatment appears;
5. an additional result can genuinely distinguish among the remaining hypotheses.

If the data remain insufficient even then, undetermined remains a legitimate expert conclusion.

[VISUAL 72.4: Escalation ladder—routine methods → advanced analysis → research collaboration → undetermined]

Chapter summary

• An advanced method is selected according to the unresolved question, not the prestige of the instrument.
• Detection limit, spatial resolution, sampling volume, and destructiveness are separate axes.
• Micro-CT provides 3D internal structure but is not a universal diamond-origin test.
• XRD primarily identifies the crystalline phase.
• XRF and LA-ICP-MS differ in sensitivity, sample volume, and destructiveness.
• Colored-stone trace-element origin models do not transfer automatically to diamonds.
• SIMS can link isotopes to microscopic growth zones.
• CL can reveal growth structures that standard deep-UV imaging sometimes does not resolve.
• EPR is a specialized method for paramagnetic defects, not a routine tester.
• X-ray topography can map a lattice-defect fingerprint.
• AI performance is valid only within the validated dataset, protocol, and version.
• Advanced analysis must include calibration, controls, and the legitimate possibility of an undetermined conclusion.