Skip to content

The Material Spectrum: From Liquids to Solids

Updated to manuscript V3

In the review the spectrum is organized by handling class (liquids, semi-solids, powders and discrete objects) rather than by a single yield-stress axis, and each class is mapped onto dispensing technologies with typical accuracy and throughput (Table 1, on Dispensing by Material Class). The yield-stress ranges below are the earlier wiki's illustrative organizing axis; for the review's sourced ranges, see Table 1. Discrete objects (vials, premix containers, lids) are treated in Object Handling.

Yield Stress as the Organizing Axis

Materials encountered in autonomous formulation laboratories span a continuum based on yield stress — the minimum stress required for flow:

┌─────────────────────────────────────────────────────────────────┐
│  10⁰ Pa         10² Pa         10⁴ Pa         10⁶ Pa         10⁹ Pa
│   └──────────────────────────────────────────────────────────────┘
│
│  LIQUIDS      SEMI-SOLIDS      POWDERS       SOLIDS
│  (flow)       (yield-stress)   (granular)    (rigid)
│
└─────────────────────────────────────────────────────────────────┘

Lower yield stress → higher mobility → continuous flow
Higher yield stress → lower mobility → discrete manipulation

Material Classes and Handling Technologies

Liquids (Yield Stress ~10⁰ – 10² Pa)

Discrete Dispensing: - Air-displacement pipettes (standard for < 100 cP) - Positive-displacement syringes - Syringe pumps

Continuous Dispensing: - Peristaltic pumps - Gear pumps

Contactless: - Acoustic droplet ejection - Piezoelectric dispensers

Dominant Strategy: Characterize (measure viscosity) → Act


Semi-Solids & Gels (Yield Stress ~10² – 10⁵ Pa)

Extrusion-Based: - Syringe-based extrusion (motor drives plunger) - Air pressure-driven (gentler on shear-sensitive bio-inks) - Screw/auger extrusion (decouples flow from pressure)

Progressive Cavity Pumps: For consistent output despite yield stress

Dominant Strategies: - Real-time correction (watch→detect→correct) - Iterative optimization (dispense→weigh→optimize) - Feedforward prediction (measure→predict→print)


Powders (Yield Stress ~10⁴ – 10⁶ Pa, Granular)

Feeding Methods: - Auger/screw feeders (volumetric control) - Vacuum-assisted dosing (suction through orifice) - Vibratory hoppers (struggles with sticky/charged powders)

Feedback: - Gravimetric (weigh continuously or at endpoints) - Vision (estimate fill height in scoops) - Tactile (touch-based retrieval)

Dominant Strategies: - Simulation-based generalization (train in sim, apply to unseen powders) - Real-time correction (scoop→verify→pour) - Iterative optimization (measure→calibrate→dispense)


Solids (Yield Stress ~10⁶ – 10⁹ Pa)

Discrete Object Handling: - 6-DOF robotic arms (UR3e, Franka, Kinova) - SCARA robots (higher precision) - Cartesian/gantry systems (lower cost)

End-Effectors: - Vacuum grippers - Parallel grippers - Soft/compliant grippers - Vision-guided pick-and-place

Dominant Strategy: Tactile retrieval (probe→feel→retrieve)


The Spectrum in Practice

The yield-stress spectrum determines not only which mechanism to use, but also which strategy to employ:

Yield Stress (Pa) Material Handling Class Primary Mechanism Primary Strategy Typical Accuracy
10⁰ – 10¹ Water, solvents Liquid Pipette/pump Characterize Excellent (< 1%)
10¹ – 10² Honey, oil Liquid Syringe/peristaltic Characterize or iterate High
10² – 10³ Thick paste, gel Semi-solid Screw extrusion Real-time correction Moderate
10³ – 10⁵ Yield-stress paste Semi-solid Air-pressure extrusion Watch→detect→correct Moderate
10⁴ – 10⁶ Powder (free-flowing) Granular Auger/vibration Simulation generation Variable
10⁶ – 10⁸ Powder (cohesive) Granular Gravimetric + vision Real-time correction Good (±1 mg)
10⁸ – 10⁹ Tablet, pellet Solid Pick-and-place Tactile retrieval Excellent

Key Insight: Strategies Recur Across Materials

While each material class presents distinct physical challenges (viscosity, flow, cohesion), the underlying control strategies are remarkably similar:

  • Both high-viscosity liquids and pastes benefit from real-time correction
  • Both powders and semi-solids use simulation-based generalization
  • All classes can employ iterative optimization when feedback sensors are available

This strategy-level view reveals that future autonomous labs should:

  1. Identify material properties (yield stress, viscosity, flowability) on the fly
  2. Select the appropriate strategy (characterize, optimize, correct, predict, simulate, reinvent, or retrieve)
  3. Apply the strategy using a shared sensing and control infrastructure
  4. Switch strategies as materials change, without re-engineering the entire system

References

  • Rastegardoost, M. M.; Bao, Z.; Sadr, S.; Abdollahi, N.; Darvish, K.; Clasky, A.; Ohsaka, H.; Le Devedec, F. Self-Driving Laboratories for Formulation: Handling Liquids, Pastes, Powders and Objects, from Robots to Digital Twins and AI Agents. Manuscript in preparation, 2026.
  • Key references from the review: [13], [14], [18], [21], [30]. Full list on the References page.

Further reading (from the earlier wiki)

  • Tadros, T. F. (2010). Rheology of Dispersions. Wiley-VCH.