Scalable Ball‑Milling Production of Nanoscale CL‑20/Graphene Oxide Composites with Reduced Particle Size and Sensitivity

Background

Unintended detonation of high explosives (HEs) such as CL‑20, HMX, and RDX remains a critical safety concern in modern conflict zones.1 The high energy density of these compounds is often offset by their poor sensitivity to impact, friction, shock, and thermal stimuli.2 Over recent decades, researchers have explored several strategies to produce insensitive energetic materials, including polymer‑bonded explosives (PBXs), microcapsule formation via in‑situ polymerisation, and energetic co‑crystals.3‑6 The coating route is the most prevalent, yet it relies heavily on organic solvents, raising environmental and safety issues.7‑9 Alternative approaches, such as co‑crystallisation of CL‑20 with HMX, have shown promise but still face challenges related to chemical compatibility and residual sensitivity.10‑15 Ball‑milling offers a scalable, solvent‑free route to produce nanocrystalline energetic materials. The process can uniformly refine particle size, narrow the size distribution, and round particle morphology—all factors that lower initiation sensitivity.16 At the same time, mechanical milling preserves the crystal structure, which is essential for maintaining the energetic performance of CL‑20.17 Graphene and its oxidised derivative, graphene oxide (GO), have attracted attention for their high thermal conductivity, mechanical strength, and ability to stabilise energetic compounds.18‑21 Prior work has demonstrated that GO can reduce the impact and shock sensitivity of HMX and RDX.8 The present study extends this concept to CL‑20, leveraging the exfoliation capacity of ball‑milling to generate few‑layer GO directly from graphite materials, thereby simplifying the production pipeline and avoiding the need for additional chemical reduction steps.

Methods

Synthesis of Nanoscale CL‑20/GO Composites

Raw CL‑20 (Liang & Co., Ltd.) was dispersed in de‑ionised water with varying weight percentages of graphite materials (GIMs): 0.5 %, 1 %, 2 %, and 5 % relative to CL‑20. The suspension was subjected to ball‑milling (zirconia beads, 0.1 mm diameter, ball‑to‑powder ratio 20, 300 RPM, 10 min). After milling, the powder was sonicated to remove any residual beads. The resulting products were designated as CL‑20, CL‑20/GO_0.5, CL‑20/GO₁, CL‑20/GO₂, and CL‑20/GO₅ (analogous nomenclature applies to reduced GO, rGO).

Characterization

Field‑emission SEM (MIRA3 LMH, Tescan) at 10 kV provided morphological data. Powder XRD (DX‑2700, Dandong Haoyuan) with Cu‑Kα radiation (λ = 1.5418 Å) covered 5°–50° (step 0.03°, 6 s). Differential scanning calorimetry (DSC) (DSX‑131, Setaram) was performed at 5, 10, and 20 °C min⁻¹. Impact sensitivity was measured using a home‑built 12‑drop hammer apparatus; the special height H₅₀ was determined from 25 trials.

Results and Discussion

SEM images (Fig. 1) show that raw CL‑20 adopts a spindle shape with an average size of ~300 µm. After ball‑milling, the grains become nearly spherical with a reduced size of ~200 nm (Fig. 1c–f). When GO is present, thin wrinkled sheets are observed coating the CL‑20 surface (Fig. 1d–g), indicating successful deposition during milling. rGO, however, shows less pronounced layering, consistent with its lower functional‑group density. XRD patterns (Fig. 2) confirm the retention of the ε‑CL‑20 crystal structure post‑milling; the characteristic peaks at 12.59°, 13.82°, and 30.29° (planes (1,1,−1), (2,0,0), and (2,0,−3)) remain, although peak intensities shift due to preferred orientation induced by milling. A distinct GO peak at 10° (0,0,2) appears in CL‑20/GO₅, corroborating the SEM evidence. The proposed formation mechanism (Fig. 3) involves simultaneous exfoliation of graphite into few‑layer GO and refinement of CL‑20 particles, leading to intercalated composites via non‑covalent interactions (hydrogen bonding) between GO functional groups and CL‑20 nitrogen atoms. rGO lacks sufficient functional groups, resulting in weaker interfacial bonding. DSC analysis (Fig. 4) reveals that nanoscale CL‑20 and its GO composites exhibit smoother, non‑truncated endotherms compared to raw CL‑20, indicating mitigated thermal runaway. Kissinger plots (Fig. 5a) and kinetic compensation analysis (Fig. 5b) show that the apparent activation energy (E_a) of CL‑20/GO₂ and CL‑20/GO₅ is slightly higher than that of other samples, yet the pre‑exponential factor (A) follows the expected linear relationship with E_a, suggesting a consistent decomposition mechanism across all specimens. Thermodynamic parameters (ΔG^≠, ΔH^≠, ΔS^≠) calculated from DSC data confirm that all composites retain the inherent stability of CL‑20; peak temperatures (T_p0) and critical explosion temperatures (T_b) remain unchanged relative to raw CL‑20. Impact‑sensitivity measurements (Fig. 6) demonstrate a progressive increase in H₅₀ with higher GO loading: CL‑20/GO₅ reaches >150 cm, whereas raw CL‑20 is ~60 cm. rGO composites exhibit lower H₅₀ values (e.g., CL‑20/rGO₅ ~120 cm) due to reduced functional‑group interaction. The enhanced lubrication and heat dissipation provided by GO layers reduce the formation of hot spots and dislocation networks during impact, thereby decreasing sensitivity. Collectively, these results confirm that ball‑milling produces nanoscale CL‑20/GO composites that are both morphologically refined and mechanically desensitised, without compromising thermal performance.

Conclusions

The scalable ball‑milling approach described herein yields CL‑20/graphene oxide composites with a ~200 nm grain size, spherical morphology, equal thermal stability, and markedly reduced impact sensitivity. The process leverages GO’s oxygen functional groups to form robust, non‑covalent interfaces that suppress aggregation and promote efficient exfoliation. This method eliminates the need for separate chemical reduction steps and is readily adaptable to other energetic systems (e.g., metal or polymer‑loaded GO composites). The resulting material is a promising candidate for booster or propellant applications where safety and performance must be balanced.

Abbreviations

2D:

Two‑dimensional

CL‑20:

2,4,6,8,10,12‑hexanitro‑2,4,6,8,10,12‑hexaazaisowurtzitane

E_a:

Apparent activation energy

EPDM:

Ethylene‑propylene‑diene monomer

FESEM:

Field‑emission scanning electron microscopy

GEMs:

Graphene materials

GIMs:

Graphite materials

GO:

Graphene oxide

H₅₀:

Special height

HEs:

High explosives

HMX:

1,3,5,7‑teranitro‑1,3,5,7‑tetrazocine

NC:

Nitrocellulose

PBXs:

Polymer‑bonded explosives

PDF:

Portable document format

RDX:

Hexahydro‑1,3,5‑trinitro‑1,3,5‑triazine

rGO:

Reduced graphene oxide/graphene

ΔG^≠:

Free energy of activation

ΔH^≠:

Enthalpy of activation

ΔS^≠:

Entropy of activation