A Practical Identification Workflow

Professional diamond identification is not about using as many instruments as possible. A good laboratory does not begin by asking “what can we measure?” but by asking “what claim must we verify?” Only then does it select the procedure that provides the most information with the least risk to the sample.

This is particularly important with diamonds because different questions may overlap in one object: whether the material is diamond at all, whether it is natural or laboratory-grown, whether it has been treated, whether its color is natural, whether the stone matches its report, and what its quality is. These questions are not the same and are not resolved by the same evidence.

Four levels of questions

A practical workflow is most stable when it follows this hierarchy:

1. material identity — diamond or another material;
2. diamond origin — natural or laboratory-grown;
3. treatment and origin of color — whether appearance was subsequently altered and, for colored diamonds, whether the color is natural or treated;
4. quality and documentary matching — grading, condition, report, inscription, and other identification layers.

Provenance, ownership, market value, and legality are separate domains. A spectrum may help identify material or defect history, but it cannot prove ownership. A laser inscription may help match a stone to a document, but by itself it proves neither natural origin nor the authenticity of the entire documentary chain.

[VISUAL 66.1: Identification hierarchy—material → origin → treatment/color origin → quality/documentation]

Intake: describe the sample first

Before measurement, record what was actually received:

• a loose or mounted stone;
• approximate mass or dimensions;
• shape and facet arrangement;
• visible color and transparency;
• number of stones in jewelry or melee;
• surface condition;
• an existing inscription, document, or seller’s claim;
• reason for examination.

This step is not an administrative formality. A contaminated surface, setting metal, a very small stone, damage, a coating, or a difficult-to-access area may change both the choice of test and the reliability of its result.

Nondestructive methods as the default rule

Nondestructive methods take priority with gem-quality diamonds. Destructive or invasive sampling is justified only when the scientific question requires it, when it is permitted, and when the consequences are clearly documented.

In routine identification, this means beginning with observation, an appropriate property test, microscopy, and nondestructive spectroscopy. More advanced methods are added when the preceding steps cannot distinguish among realistic alternatives.

Appearance-first is not appearance-only

Visual examination provides orientation. It may reveal:

• doubled facet junctions;
• characteristic inclusions;
• growth zoning;
• a surface-reaching fracture;
• a laser-drill channel;
• coating wear;
• unusual luminescence;
• a strain pattern;
• traces of processing or contamination.

An observational sign has evidentiary value only in context. “There are no inclusions,” “metal is visible,” “it has blue fluorescence,” or “it looks like CVD” are not sufficient conclusions about origin.

Screening is not confirmation

Screening is the process of separating samples that meet a predefined criterion from those that require further testing. A refer result is therefore not synonymous with “laboratory-grown,” “treated,” or “fake.” It means that the device or protocol cannot safely complete the case within its own scope.

This is especially important with modern diamond-screening systems. Current GIA literature emphasizes that no single fast, inexpensive screening device definitively identifies every laboratory-grown diamond across all sizes, colors, shapes, and treatment conditions. Final identification in more demanding cases rests on a combination of gemological and analytical findings.

Loose, mounted, and melee are not the same problem

A loose stone provides the best access to the girdle, pavilion, culet, and different optical paths. A mounted stone may have concealed areas, metallic artifacts, and limited probe contact. Melee adds the problem of large sample numbers, very small dimensions, and the need for high-throughput screening.

The result must therefore always be interpreted according to the sample actually tested, not an ideal laboratory scenario.

Select the method according to the question

After the initial examination, select the method with the greatest discriminatory power for the specific question.

• Handheld property tester: rapid differentiation of certain materials; does not resolve natural/laboratory-grown origin.
• Microscopy: morphology, position, surface relationships, growth and treatment clues, and documentation.
• UV-Vis-NIR absorption: optical centers associated with color and origin-of-color problems.
• FTIR: diamond type, nitrogen aggregation, and other bulk impurity and defect data.
• Raman: phase identity and localized identification of an inclusion or surface material.
• PL: highly sensitive detection of certain defect centers, often with controlled temperature and excitation wavelength.
• Luminescence imaging: spatial growth and defect patterns; covered in detail in Chapter 71.
• Advanced methods: used when routine methods do not provide sufficient evidentiary resolution; Chapter 72.

[VISUAL 66.2: Map of methods by question—material identity, origin, treatment, and color origin]

Contradictory results are information

If one finding suggests one explanation and another suggests something else, the professional response is not to “believe the more expensive instrument.” Instead, verify:

1. whether the methods measured the same property;
2. whether one method was outside its validated scope;
3. whether an artifact arose from the sample, setting, surface, or instrument;
4. whether a realistic material or treatment scenario explains both results;
5. which additional test best distinguishes the remaining hypotheses.

A contradiction often reveals that the initial question was too broad or that a result was misinterpreted.

Levels of conclusion

It is useful to distinguish among the following formulations:

• confirmed — the body of evidence is sufficient for the specific conclusion;
• compatible with — observations fit the hypothesis but are not unique to it;
• refer — screening did not complete the identification;
• further testing required — the question is clear, but the current dataset is insufficient;
• undetermined — the available methods cannot reliably distinguish the relevant alternatives.

“Undetermined” is not a professional failure when a genuine evidentiary boundary exists.

Chain of evidence

Every serious conclusion should be reconstructible. The minimum record includes:

• sample identity and condition;
• date;
• method used and relevant conditions;
• direct observations or measurement results;
• interpretation;
• alternative explanations excluded or left open;
• final level of confidence.

For high-value, disputed, or legally sensitive objects, the level of documentation must increase with the potential consequences of an error.

A unified practical workflow

Step 1 — define the question. What must be demonstrated, and what falls outside the test?

Step 2 — document the sample. Loose, mounted, or melee; dimensions, appearance, condition, and existing documents.

Step 3 — remove reversible interference. Safe cleaning and stabilization of conditions when appropriate.

Step 4 — perform an orienting visual and microscopic examination. Record the observation before interpreting it.

Step 5 — confirm material identity. Do not proceed to origin until the material itself is known.

Step 6 — if it is diamond, decide whether natural/laboratory-grown screening is required. Read the result within the method’s declared scope.

Step 7 — if warranted, test treatment or color origin. Select the method according to the type of problem.

Step 8 — compare all data. Look especially for contradictions.

Step 9 — escalate only when a new test can change the conclusion. More measurements do not automatically create more knowledge.

Step 10 — formulate the conclusion within the limits of the evidence. Write identity, origin, treatment, and quality as separate axes.

[VISUAL 66.3: Decision tree 66—from intake to a confirmed/refer/undetermined conclusion]

The most common errors

The most dangerous shortcuts are:

• “the tester says diamond, so it is natural”;
• “refer means laboratory-grown”;
• “there are no inclusions, so it is CVD”;
• “a metallic inclusion proves HPHT”;
• “one peak resolves origin”;
• “a laser inscription proves that the stone is the one shown in the report”;
• “a more expensive instrument always outranks a cheaper one”;
• “if we run enough tests, the conclusion must become certain.”

These are not merely technical errors. They all result from conflating measurement, interpretation, and conclusion.

[VISUAL 66.4: Pyramid of evidentiary strength—observation → measurement → interpretation → integrated conclusion]

Chapter summary

• Identification begins with a question, not an instrument.
• Material identity comes before natural/laboratory-grown origin.
• Treatment status and origin of color are separate from material origin.
• Provenance, ownership, and value cannot be read from a spectrum.
• Intake and sample condition are part of the analytical procedure.
• Nondestructive methods take priority in routine gemology.
• An appearance-first examination provides direction but rarely completes a complex case by itself.
• Screening and confirmation are not synonyms.
• Refer does not mean laboratory-grown or treated.
• Loose, mounted, and melee samples require different practical protocols.
• Contradictory results require explanation, not an arbitrary choice of the “better” device.
• A conclusion must be stated at the level of confidence that the actual evidence permits.