Dunhuang Murals at the Micron Scale: Pigment Identification and Layer Thickness Mapping by Micro-LIBS

Technical Summary: Dunhuang Murals at the Micron Scale: Pigment Identification and Layer Thickness Mapping by Micro-LIBS

Problem Statement
Mural paintings are fragile cultural heritage assets severely compromised by deterioration processes such as chalking, flaking, and salt efflorescence. Effective conservation requires precise characterization of pigment composition, particle size distribution, and layer stratigraphy. However, conventional analytical methods face a dichotomy: invasive techniques cause irreversible damage to inherently fragile murals, while non-invasive methods (e.g., XRF, FTIR, Raman) often suffer from signal degradation due to complex substrates and surface contamination, limiting their efficacy in stratigraphic analysis. There is a critical need for minimally invasive analytical techniques that offer high measurement accuracy and spatial resolution without compromising the artistic integrity of the relics.

Methodology
The study employs Micro-Laser-Induced Breakdown Spectroscopy (Micro-LIBS) to analyze a detached mural fragment (MG-01) from the Dunhuang Mogao Grottoes. The experimental approach involved three primary phases:

1. Parameter Optimization: To achieve a "nearly non-destructive" analysis, the authors systematically optimized Micro-LIBS parameters using simulated mural samples.

   • Delay Time: Optimized to 1.2 μs to maximize the signal-to-background ratio (SBR) and minimize relative standard deviation (RSD) by allowing the plasma continuum to decay.
   • Focal Position: Determined that setting the laser focal point 0.25 mm below the target surface (rather than the conventional 2 mm) minimized ablation crater diameter and pigment discoloration while maintaining spectral intensity.
   • Laser Energy: Selected 2.0 mJ per pulse. While lower energies (0.5 mJ) prevented discoloration, they yielded poor spectral stability. Higher energies (>2.5 mJ) caused non-linear ablation and excessive thermal damage (oxidation of red lead). The 2.0 mJ setting balanced spectral quality with minimal sample damage.

2. Sample Analysis: Using the optimized parameters (2.0 mJ, 1.2 μs delay, 0.25 mm defocus), the system performed point-by-point laser scanning on the actual mural fragment.

   • Pigment Identification: LIBS spectra were analyzed for characteristic elemental lines (Cu, Pb, Ca, etc.) and cross-validated with Raman spectroscopy.
   • Stratigraphic Mapping: The evolution of emission line intensities (Cu I, Pb I, Ca II) was monitored as a function of consecutive laser pulses. A calibration model relating laser pulse count to ablation depth was established using polynomial fitting of crater depth data from simulated samples.

3. Validation: Results were verified against cross-sectional optical microscopy to confirm layer thickness and painting sequence.

Key Results

• Pigment Identification: The study successfully identified the green pigment as atacamite (Cu₂(OH)₃Cl), the red pigment as red lead (Pb₃O₄), and the white ground layer as anhydrite (CaSO₄). These identifications were confirmed by matching LIBS characteristic lines with Raman spectral peaks (e.g., Cu-Cl stretching at 363 cm⁻¹ for atacamite; SO₄²⁻ symmetric stretching at 1018 cm⁻¹ for anhydrite).
• Layer Thickness and Sequence: By monitoring the intensity shifts of elemental lines during sequential ablation, the authors reconstructed the painting sequence: Ground Layer → Red Pigment → Green Pigment.
  • At the green pigment site (P1), the Cu signal dominated initially, while the Pb signal (from the underlying red layer) emerged and intensified after the 3rd laser pulse, indicating the complete removal of the green layer.
  • At the red pigment site (P2), the Pb signal weakened and Ca signals (from the ground layer) strengthened after the 3rd pulse.
  • Using the pulse-depth calibration, the green layer thickness was estimated at ~35.5 μm and the red layer at ~38.5 μm. These values aligned closely with cross-sectional microscopy measurements (35.51 ± 3.67 μm and 38.51 ± 4.42 μm, respectively).
• Damage Assessment: Under the optimized conditions, ablation craters were approximately 46.86 μm in diameter with a discolored area of ~138.89 μm. The damage was described as "naked-eye undetectable" and significantly smaller than that produced by conventional LIBS.

Significance and Claims
The paper claims that Micro-LIBS, when systematically optimized, serves as a reliable, high-spatial-resolution, and near-non-destructive tool for the stratigraphic analysis of ancient murals. Key contributions include:

• Resolution of Stratigraphy: The method successfully resolves individual pigment layers in polychrome murals, a task where conventional non-invasive methods often fail due to signal interference.
• Painting Sequence Reconstruction: The ability to monitor layer-by-layer ablation allows for the determination of the painting sequence (ground-red-green) without the need for invasive cross-sectioning.
• Quantitative Thickness Mapping: The study demonstrates that Micro-LIBS can quantify individual layer thicknesses with sub-5 μm accuracy, providing data essential for authentic restoration efforts.
• Minimally Invasive Alternative: The work positions Micro-LIBS as a pragmatic alternative for scientific analysis, bridging the gap between fully non-destructive techniques (which lack depth resolution) and invasive methods (which cause irreversible harm).

The authors conclude that this approach provides critical technical support for cultural heritage conservation, particularly for elucidating fabrication procedures and painting techniques of the Dunhuang Mogao Grottoes, and suggest future applications for in-situ profiling of salt efflorescence and deterioration products.