Evaluation of Particle Size in Chocolate —Evaluation of Particle Size Distribution and Coarse Particles that Affect Texture

Significance of the Topic

Particle size distribution (PSD) and the abundance of coarse particles are critical quality attributes for multiphase food products such as chocolate because they strongly influence mouthfeel, perceived smoothness, and overall texture. Quantifying both the volumetric PSD and the number concentration of particles above threshold sizes supports process control, formulation optimization, and product release criteria in food manufacturing.

Objectives and Study Overview

The study compares two complementary measurement approaches for solid particles in milk chocolate: laser diffraction (SALD-2300) for volumetric PSD and a dynamic particle image analysis system (iSpect DIA-10) for particle sizing and number concentration in the coarse-particle range. Two commercial milk-chocolate samples (F and UF), previously assessed by informal sensory panels with UF judged smoother than F, were analyzed to link instrumental metrics to perceived texture.

Methodology

Sample preparation and dispersion:

• Approximately 0.1 g of thinly sliced chocolate was scraped into a 50 mL beaker.
• 20 mL of isopropanol heated to ~40 °C was added to dissolve fats and disperse solid particles.
• The suspension was sonicated in a 240 W ultrasonic bath for 5 minutes.

Notes on dilution and measurement scope:

• For SALD-2300 measurements the dispersed suspension was used in a stirred batch cell at an appropriate concentration for laser diffraction.
• For iSpect DIA-10, an additional 50× dilution was performed, resulting in a final equivalent concentration of 0.1 g chocolate per 1000 mL isopropanol; this ensured single-particle detection conditions for image analysis.

Used Instrumentation

• SALD-2300 Laser Diffraction Particle Size Analyzer — batch cell with stirring; refractive index used for solids: 1.70 − 0.02i.
• iSpect DIA-10 Dynamic Particle Image Analysis System — image acquisition at 8 fps (shooting efficiency 97 %), binarization threshold 110, sampled liquid volume per measurement 0.1 mL.

Main Results and Discussion

Volumetric particle size (SALD-2300):

• Both samples showed a lower PSD limit near ~0.7 µm. The upper tail differed markedly: sample F extended to about 90 µm, while sample UF extended to about 40 µm.
• Percentage of particles above 65 µm (volumetric basis): sample F = 1.77 %; sample UF = 0.00 % — indicating a higher coarse-particle content in F, consistent with sensory impressions of roughness.

Number concentration and image-based sizing (iSpect DIA-10):

• The iSpect system cannot detect particles below ~5 µm; SALD indicated ~20–30 % of particles lie below this threshold, so iSpect measures only a portion of the total PSD but captures the coarse-particle population reliably.
• Total particle counts detected: sample F = 14,746 particles; sample UF = 6,861 particles (coarser-particle region resolved more densely in F).
• Measured number concentrations (particles/mL) and corresponding solid-particle contents (converted to particles per gram of chocolate):
  
  • ≥20 µm: F = 308 particles/mL (≈3.08 × 10^6 particles/g); UF = 128 particles/mL (≈1.28 × 10^6 particles/g)
  • ≥40 µm: F = 36 particles/mL (≈3.6 × 10^5 particles/g); UF = 5 particles/mL (≈5.0 × 10^4 particles/g)
  • ≥50 µm: F = 19 particles/mL (≈1.9 × 10^5 particles/g); UF = 0 particles/mL
• Scatter-plot analyses of maximum particle length vs. aspect ratio revealed a noticeably higher frequency of large particles (>50 µm) in sample F than in UF.

Interpretation relative to texture:

• Instrumental data from both methods consistently indicate sample F contains more and larger coarse particles, which matches sensory panel reports that F felt slightly rougher.
• Perceived roughness is multifactorial: particle size/number are important but interact with fat/oil content and melting behavior of the matrix; therefore combined measurement approaches give a fuller picture.

Benefits and Practical Applications

• Combining volumetric PSD (laser diffraction) with image-based number-concentration measurements provides complementary information: laser diffraction quantifies fine-mode distribution (including <5 µm population) while image analysis delivers exact counts and sizes in the coarse-particle region that drive mouthfeel.
• These methods can be applied in routine QC to monitor milling/conching efficacy, to set pass/fail criteria based on coarse-particle count thresholds, and to guide formulation changes (e.g., fat content, emulsifiers) to improve smoothness.
• Number-concentration metrics (particles per g above specified size) are particularly useful as actionable quality indicators for sensory-related defects.

Future Trends and Potential Applications

• Integration of multimodal particle characterization (laser diffraction, high-resolution imaging, and possibly single-particle light scattering) will improve resolution across the full PSD and reduce blind spots below instrument detection limits.
• Automated image analysis with machine learning could classify particle morphology and composition (e.g., sugar crystals vs. cocoa solids) to better predict sensory outcomes.
• Standardization of coarse-particle number thresholds linked to sensory panels would enable harmonized acceptance criteria across manufacturers.
• These approaches can be extended beyond chocolate to other food systems, pharmaceuticals, and materials where low concentrations of coarse particles critically affect performance or perception.

Conclusion

Using SALD-2300 and iSpect DIA-10 in tandem allows robust evaluation of chocolate particle populations: laser diffraction quantifies overall PSD including fine fractions, and image analysis accurately enumerates coarse particles that correlate with perceived roughness. The complementary dataset supports more targeted process control and product optimization than either method alone. Limitations include the image analyzer’s lower-size cutoff (~5 µm) and the influence of matrix properties (fat content, melting) on sensory perception; addressing these through combined methods and sensory calibration improves predictive power.

References

• Kinoshita T. Evaluation of Particle Size in Chocolate — Evaluation of Particle Size Distribution and Coarse Particles that Affect Texture. Shimadzu Application News, First Edition July 2026. Document number 01-00624-EN.