Enhancing CL‑20 Safety: Lowering Impact Sensitivity via Carbon Nanomaterial‑Assisted Thermal Conductivity

Background

CL‑20 (2,4,6,8,10,12‑hexanitro‑2,4,6,8,10,12‑hexaazaisowurtzitane) offers superior density and energy density compared with traditional explosives such as RDX and HMX. Its high sensitivity, however, stems from poor thermal conductivity, which promotes the formation of hot spots during rapid temperature fluctuations and jeopardizes weapon‑system reliability.1–7 Addressing this limitation is therefore critical for safe, high‑performance applications.

Polymer coatings and graphite fillers have long been employed to improve the mechanical and thermal robustness of energetic crystals. Recent consensus favors the use of carbon‑based nanofillers—graphene nanoplatelets, CNTs, and related materials—to create efficient thermal pathways in polymer composites.8–13 These nanomaterials possess exceptional in‑plane thermal conductivity and high aspect ratios, enabling the formation of continuous, low‑resistance heat‑transfer networks.

While graphene’s two‑dimensional lattice offers a large phonon mean free path and high electron mobility, interlayer van der Waals interactions often impede cross‑plane heat flow. CNTs, by contrast, provide one‑dimensional bridges that can link graphene sheets and explosive particles, reducing interfacial thermal resistance and forming a three‑dimensional network.14–16 Thus, a synergistic blend of rGO and CNT is anticipated to deliver superior thermal conductivity and, consequently, reduced impact sensitivity.

This work explores the combined effect of rGO and CNT in CL‑20 composites, characterizing microstructure, thermal behavior, sensitivity, and detonation performance, and elucidating the underlying heat‑transfer mechanisms.

Methods

Synthesis of Nanoscale CL‑20/Carbon Material Composites

CL‑20 composites were fabricated via a water‑suspension route (Fig. 1). Estane (3 wt % in 1,2‑dichloroethane) served as the binder matrix. rGO, CNT, or a rGO:CNT 2:1 mixture (SWCNT) were dispersed in the binder solution by ultrasonic agitation. Separately, 20 g milled CL‑20 was suspended in 200 mL de‑ionized water under magnetic stirring. The binder solution was slowly added to the CL‑20 suspension, and the mixture was heated at 70 °C under 0.02 MPa pressure to remove solvent. After cooling, the composite was filtered, washed, and vacuum‑evaporated to yield the final product. Samples were labeled as CL‑20/Estane (1), CL‑20/rGO (2), CL‑20/CNT (3), and CL‑20/rGO + CNT (4).

Experimental diagram of CL‑20‑based composites prepared by water‑suspension method

Characterization

SEM (SU‑8020, Hitachi) captured surface morphology and particle size distribution. XRD (DX‑2700, Dan Dong Hao Yuan) assessed crystal structure using Cu‑Kα radiation (40 kV, 30 mA). DSC (DSC‑131, Setaram) evaluated thermal decomposition under a nitrogen flow (30 mL min⁻¹) with heating rates of 5–20 K min⁻¹. Static‑electricity accumulation was measured by inducing friction on a chute and recording charge with a digital meter; the charge per unit mass quantified electrostatic buildup. Impact sensitivity followed GJB 772A‑97 (type 12 drop hammer) to determine the 50 % probability height (H₅₀). Thermal diffusivity was obtained via laser‑flash analysis on 10 mm × 2 mm discs, and thermal conductivity (k) calculated from the standard formula. Detonation velocity was measured with a time‑of‑flight probe, and theoretical values were computed using EXPLO5.

Results and Discussion

Microstructure Characteristics

SEM images (Fig. 2) show raw CL‑20 as ~300 µm spindle particles (2a). Ball‑milling reduced the size to ~200 nm (2b). rGO sheets (~2 µm, five layers) and CNTs formed interlocking assemblies (2c). In CL‑20/CNT composites, CNTs tended to agglomerate, impairing heat transfer (2d, e). Conversely, the rGO + CNT blend displayed uniform dispersion with no discernible agglomerates (2f), indicating effective synergy.

SEM morphologies of CL‑20, the mixture of rGO and CNT, and CL‑20-based composites: a raw CL‑20; b milled CL‑20; c rGO + CNT; d, e CL‑20/CNT; and f CL‑20/rGO + CNT

XRD patterns (Fig. 3) confirm the ε‑phase of all samples, with peak positions unchanged but reduced intensities and broadened widths due to the nanofiller coating and particle size effects.

X‑ray diffraction patterns of samples

Thermal Analysis

DSC curves (Fig. 4) reveal an exothermic peak at 242 °C for raw CL‑20, characteristic of explosive decomposition. Coated samples exhibit similar peak shapes but a ~2 °C shift in the rGO + CNT blend, indicating superior compatibility. Earlier decomposition onset for all composites suggests catalytic activity from the nanofillers. CNT addition reduces decomposition enthalpy from −2384.95 to −779.82 J g⁻¹, potentially diminishing explosive heat; rGO moderates this effect, yielding a balanced −1897.80 J g⁻¹. Hence, CNT content must be carefully controlled to preserve energetic output.

DSC curves of samples

Sensitivity Analysis

Impact sensitivity, quantified by H₅₀, improved dramatically from 17.3 cm (raw CL‑20) to 65.8, 50.3, and 68.7 cm for samples 2, 3, and 4, respectively. The dense, thermally conductive coating formed by rGO and CNT suppresses hot‑spot generation and dissipates mechanical energy efficiently, lowering sensitivity. Static‑electricity tests (Fig. 6) confirm reduced charge accumulation for all composites, with the rGO + CNT blend showing the lowest values, attributable to CNT’s charge‑dissipating capability.

Impact sensitivity of samples

Static electricity accumulation of samples

Thermal Conductivity Analysis

Measured at 25 °C, raw CL‑20 exhibits 0.143 W m⁻¹ K⁻¹. Coating with 1 wt % rGO + CNT increases conductivity to 0.64 W m⁻¹ K⁻¹—4.5× the baseline—while rGO or CNT alone deliver smaller gains. The high intrinsic conductivities of rGO and CNT, combined with their interfacial synergy, establish a continuous, three‑dimensional heat‑transfer network that minimizes interlayer resistance. A schematic (Fig. 7) illustrates CNTs bridging rGO sheets and CL‑20 particles, creating numerous phonon pathways and maximizing surface contact.

Schematic diagram of thermal transfer of CL‑20/rGO + CNT

Linear regression (Fig. 8) demonstrates a clear inverse relationship between thermal conductivity and impact sensitivity: as k increases, H₅₀ rises, confirming that higher thermal conductivity suppresses hot‑spot formation and mechanical ignition. The empirical fit y = 85.62527 − 101.06403 exp(−x/0.35142) captures this trend, where x is k (W m⁻¹ K⁻¹) and y is H₅₀ (cm).

Relation diagram between thermal conductivity and special height

Detonation Performances

Theoretical detonation parameters (EXPLO5) and measured velocities for raw CL‑20 and composites are summarized in Table 2. While actual velocities are slightly below theoretical predictions—attributable to environmental and instrumental factors—the inclusion of CNT alone reduces velocity by ~200 m s⁻¹, whereas the rGO + CNT blend retains performance comparable to the base material. Thus, the synergistic filler preserves energetic output while enhancing safety.

Conclusions

Incorporating rGO and CNT into CL‑20 markedly elevates thermal conductivity, forming a robust, three‑dimensional heat‑transfer network. This enhancement translates into a substantial drop in impact sensitivity, as quantified by H₅₀, without compromising detonation energy. The empirical correlation between k and sensitivity underscores the pivotal role of thermal management in explosive safety. Future work will explore optimal rGO:CNT ratios to further balance performance and safety.

Abbreviations

CFRP:

Carbon fiber reinforced plastic

CL-20:

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

CNT:

Carbon nanotube

DSC:

Differential scanning calorimeter

GNP:

Graphene nanoplatelets

H₅₀:

Special height

HMX:

1,3,5,7‑Teranitro‑1,3,5,7‑tetrazocine

PBX:

Polymer‑bonded explosive

RDX:

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

rGO:

Reduced graphene oxide

SEM:

Scanning electron microscopy

SWNT:

Single‑walled carbon nanotube

VDW:

Van der Waals force

XRD:

X‑ray diffraction