Absorption Spectroscopy

Color is not a sufficiently precise description of what is happening in a crystal. Two diamonds can look similar yet have different absorption mechanisms; two spectra can share one marker yet reflect different growth or treatment histories.

Absorption spectroscopy therefore asks not only “what color is the stone?” but which wavelengths the sample absorbs and what set of optical centers can explain that pattern.

Absorption, transmission, and observed color

When light enters a diamond, some is reflected, some scattered, and some transmitted, while particular wavelengths may be absorbed. The observed color is the result of the entire optical system, including the light source, optical path, faceting, and luminescence.

A spectrum is therefore not a “photograph of color.” It is a measurement record of the interaction between light and matter.

Handheld spectroscope and UV-Vis-NIR spectrophotometer

A handheld spectroscope enables observation of stronger lines and bands in the visible region, but the result depends on the eye, illumination, and strength of the feature.

A UV-Vis-NIR spectrophotometer provides a digital record of absorption or transmission across a broader range. It allows the positions, widths, and intensities of features to be compared, but still requires control of geometry, baseline, reference, temperature, and signal quality.

Broad bands and sharp lines carry different information

A narrow line or zero-phonon line (ZPL) may indicate a precisely defined electronic transition of a particular center. A broad absorption band may dominate the observed color even though it has no single sharp maximum.

The largest spectral feature is therefore not automatically the primary cause of color, and a marker important to identification need not strongly affect face-up appearance.

Representative diamond examples

Several frequently used examples illustrate the logic, but they do not serve as “one-peak” diagnostics:

• N3 has a ZPL at approximately 415.2 nm and is associated with the cape group of absorptions important in many natural yellow diamonds;
• H3 has a characteristic ZPL at approximately 503.2 nm and may contribute to absorption and green luminescence;
• GR1 has a ZPL at approximately 741 nm and an associated broad absorption extending roughly through the red part of the spectrum; it arises from vacancies related to natural or artificial irradiation;
• the broad 550 nm deformation band is important in a large number of natural pink–red–brown diamonds, but its atomic structure is not fully resolved.

[VISUAL 69.1: Schematic UV-Vis-NIR spectrum with the N3, H3, GR1, and 550 nm regions—educational, not an actual laboratory spectrum]

Peak assignment has a confidence level

It is useful to distinguish among:

• a firmly identified center;
• a probable assignment;
• an empirical marker whose structure is not fully known;
• an indeterminate feature.

Laboratory literature is continually evolving. It is therefore professionally dangerous to turn every peak from a reference table into a definitive story about atomic structure and history.

GR1 as an example of an incorrect shortcut

GR1 is a radiation-related vacancy center. Irradiation, however, may be natural or artificial. The presence of GR1 alone therefore does not answer the question whether the color was treated.

For color origin, the spatial distribution of color, radiation stains, the intensity and relationships of other centers, possible annealing products, and other spectroscopic and gemological data are also considered.

Temperature changes the spectrum

Low temperatures can narrow and resolve spectral features that are broad or weak at room temperature. In part of its color-origin work, GIA uses Vis-NIR spectra at approximately 77 K to obtain a clearer view of relevant optical centers.

But “low temperature is better” is not a universal rule. The question determines whether this condition is needed, and temperature must be part of the record because spectra collected at different temperatures cannot be compared uncritically.

Optical path and Beer–Lambert logic

In an ideal homogeneous sample, absorption depends on the concentration of the absorbing species and the length of the optical path. A faceted diamond is not an ideal parallel plate: facets reflect light, the path may be multiple, and the geometry changes with position.

The Beer–Lambert relationship is therefore used in gemology as physical logic, not as justification for false quantitative precision on every faceted stone.

A deeper or larger stone may show stronger total absorption because of a longer optical path, but its face-up color also depends on cut geometry.

Spectrum quality

Before interpretation, check:

• signal-to-noise ratio;
• baseline;
• reference measurement;
• detector saturation;
• insufficient absorption;
• stray light;
• scattering and reflection;
• orientation and possible polarization dependence;
• limitations imposed by a setting.

“Feature not detected” means only that it was not detected under the given conditions and above the method’s sensitivity limit. It does not necessarily mean that the feature is physically absent.

[VISUAL 69.2: Good and poor spectra—baseline, noise, saturation, and stray-light artifacts]

A color-causing feature is not the same as a diagnostic marker

A spectral feature may be:

• the primary cause of color;
• a secondary contribution to color;
• a marker of treatment history;
• a marker of growth chemistry;
• entirely unimportant to visible appearance, yet important to identification.

This distinction is particularly important for fancy-color diamonds. Color grade is a perceptual result of standardized grading; spectroscopy explains the physics and helps determine color origin, but it does not replace the grading procedure.

Color origin is a multimethod problem

UV-Vis-NIR can be a key method, but it is not sufficient for every case. Similar defect centers may occur in natural, treated, and laboratory-grown contexts.

In complex cases, the spectrum is integrated with:

• FTIR diamond type and impurity information;
• PL centers;
• microscopy;
• luminescence growth patterns;
• established treatment physics.

A practical workflow for reading a spectrum

1. Define whether the record represents absorbance, transmission, or another quantity.
2. Check the horizontal and vertical axes.
3. Record the temperature and geometry.
4. Assess the baseline, noise, and saturation.
5. Identify broad bands.
6. Identify sharp lines/ZPLs.
7. Separate color-causing features from diagnostic markers.
8. Assign a confidence level to each attribution.
9. Consider optical-path and orientation effects.
10. Propose at least one alternative interpretation.
11. Select the next method that best distinguishes among the alternatives.
12. State the conclusion within the limits of the data.

[VISUAL 69.3: Spectrum-reading workflow in 12 steps]

Three short examples

Yellow diamond. Cape features can strongly support a natural nitrogen-related color mechanism, but it is not appropriate to attribute every yellow stone automatically to N3.

Green diamond. GR1 confirms radiation-related vacancies, but alone it does not distinguish natural from artificial irradiation.

Pink diamond. A broad 550 nm band may be dominant in a natural pink–red system, but the origin conclusion still includes other spectroscopic and structural data.

[VISUAL 69.4: Three different color-origin scenarios—yellow, green, and pink—with the limitations highlighted]

Chapter summary

• An absorption spectrum is not a photograph of observed color.
• A handheld spectroscope and a UV-Vis-NIR spectrophotometer have different resolution and documentation capabilities.
• A broad band and a sharp/ZPL feature are not the same kind of information.
• The N3, H3, GR1, and 550 nm systems are useful representative examples, not universal “codes.”
• GR1 indicates a radiation-related vacancy but alone does not distinguish natural from artificial irradiation.
• Peak assignments may have different confidence levels.
• Low temperature can reveal or sharpen features, but the conditions must be recorded.
• Optical path affects absorption; a faceted stone is not an ideal Beer–Lambert sample.
• Baseline, noise, saturation, and stray light must be checked before interpretation.
• A nondetected feature is not the same as proof of its absence.
• A color-causing feature and a diagnostic marker are not synonyms.
• Color origin in complex cases requires a multimethod body of evidence.