**Objective**  
Classify the provided list of candidates into **NO MORE THAN THREE groups** according to the **[Target_Property_Descriptor]**, or place them all in **ONE group** if the data justify no split.  
Your primary evidence must come from **quantitative, database-level descriptors** (experimental or computed) that are routinely reported for this item type.

**Crucial Instructions**  
1. **Prioritise Quantitative Data**  
   - Use measurable values (e.g., band-gap in eV, toughness in J m⁻³, π→π* gap in nm, redox potential in V, dielectric constant, Young’s modulus, H-bond donor count, etc.).  
   - Cite the numeric range or threshold you rely on (even if you simulate the lookup).

2. **Minimise Intuition**  
   - Do **not** use informal, qualitative knowledge such as “looks polar” or “commonly used as insulator”.  
   - If no direct value exists, derive a proxy that **could** be tabulated (e.g., fraction of heavy atoms, aspect ratio, phonon frequency cutoff) and state the proxy.

3. **Adhere to Requested Categories**  
   - Map each item to the **exact categories** supplied below.  
   - Flag borderline cases and explain which numeric value(s) create the ambiguity.

**Candidate Items to Cluster**  
t : [0.7, 1.05, 1.4]
