Catalytic ignition of the [BMIM]DCA-H₂O₂ propellant with the Cu(vim)₂(DCA)₂ complex

Abstract

An ionic liquid (IL)-H₂O₂ propellant offers a green alternative to the traditional hydrazine-N₂O₄ system. Due to the weak oxidizing ability of H₂O₂, typical ILs cannot be spontaneously ignited. To address this issue, catalytic ignition was proposed, and a series of metal complex catalysts with the formula M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) were synthesized. Single-crystal X-ray diffraction analysis shows that the Mn, Fe and Co complexes adopt an octahedral geometry, while the Cu complex exhibits a planar quadrilateral geometry. The physicochemical properties (T_d, ΔH_c, ΔH_f, and I_sp) of the fuels (10 wt% M(vim)₂(DCA)₂ + 90 wt% [BMIM]DCA) were evaluated. Dynamic light scattering and UV-vis spectroscopy analyses confirm that the four complexes disperse well in [BMIM]DCA. Ignition tests reveal that Cu(vim)₂(DCA)₂ exhibits the best catalytic performance for the ignition of [BMIM]DCA with H₂O₂, achieving t_id as short as 21 ms. EPR studies indicate that Cu(vim)₂(DCA)₂ catalyzes H₂O₂ to generate reactive oxygen species (·OH, ·O₂⁻, and ¹O₂), which subsequently induce the spontaneous ignition of [BMIM]DCA. DFT calculations reveal that the zero-dimensional geometry of Cu(vim)₂(DCA)₂ features an unsaturated Cu center and electron-rich ligand framework. This typical structure enables efficient electron transfer to activate H₂O₂, boosting ROS generation and ignition. This work clarifies the crucial role of the metal center in the IL-H₂O₂ system and provides an important insight into the catalytic ignition of the “double-green” propellants.

---

Introduction

Propellants serve as the power source for spacecraft and are crucial for deep-space exploration. Conventional hydrazine-N₂O₄ propellants offer high energy density and specific impulse.^1,2 However, they are flammable, highly carcinogenic, and corrosive during storage, transport, and operation.^3–5 In 1990, the International Green Propellant Working Group was established to guide the development of sustainable propellants.⁶ Combining ionic liquids (ILs) with hydrogen peroxide (H₂O₂) represents a promising candidate for double-green propellants. ILs possess low volatility and high structural tunability,^7–9 while H₂O₂ acts as a clean oxidizer.^10,11 The IL-H₂O₂ propellants meet the key requirements for sustainable propulsion systems. This combination allows precise control of the fuel-oxidizer ratio and provides stable thrust.^12,13

The IL-H₂O₂ propellants align with green and sustainable principles but face challenges such as long ignition delay times (t_id) or even non-hypergolic behavior, mainly due to the limited oxidizing ability of H₂O₂. The catalytic ignition concept of ionic liquids with HNO₃ for propellant applications was first demonstrated in 2010,¹⁴ establishing the foundation for using catalysts to trigger IL combustion. Introducing catalysts to enhance the H₂O₂ decomposition to strongly oxidizing species, such as ·OH, ·O₂⁻, and ¹O₂, is the key to overcome the ignition problem.^15–18 In recent studies, the t_id of ILs-H₂O₂ has been found to be still too long for the catalytic ignition of propellants, such as [N(CH₃)₄]_n(Cu₂I₃)_n and [FcCH₂N(CH₃)₃]₂(Cu₂I₄), which give t_id values of 55 ms and 54 ms for [EMIM][BH₃CN]-H₂O₂, respectively.¹⁹ The fixed electronic environment and strong ionic bonding in these compounds slow down the dissociation to release active radicals.^20–22 Moreover, their high-dimensional structures introduce significant steric hindrance that blocks active sites.^23–25 Therefore, more efficient catalysts are required to address the ignition challenge of the IL-H₂O₂ propellants.

Previous work has demonstrated that zero-dimensional complexes can significantly reduce the ignition delay time of [BMIM]DCA IL-H₂O₂ (t_id = 30 ms, Fig. 1). Low-dimensional metal complexes exhibit prolonged stable dispersion in ILs while facilitating the efficient decomposition of H₂O₂, thereby producing highly reactive oxygen species (ROS). The mechanism originates in the optimal d-orbital (metal)–LUMO (H₂O₂) overlap, highly reducing the O–O bond cleavage energy.^26–28 Exposing the active sites through the design of coordination-unsaturated, low-dimensional complexes further enhances the electron transfer during H₂O₂ activation, achieving faster catalytic ignition.^29,30

Fig. 1 Catalytic ignition of the [BMIM]DCA-H₂O₂ propellant.

In this work, a series of metal complexes, M(vim)₂(DCA)₂ (M = Mn/Fe/Co/Cu, vim = vinylimidazole, and DCA = N(CN)₂), were synthesized. Cu(vim)₂(DCA)₂ exhibited the best catalytic performance with a minimum t_id value of 21 ms. Dynamic light scattering and UV-vis spectroscopy analyses confirmed the long-term stability of Cu(vim)₂(DCA)₂ dispersion in [BMIM]DCA. EPR spectroscopy revealed that H₂O₂ was catalyzed to generate ROS (·OH, ·O₂⁻ and ¹O₂), which subsequently triggered rapid ignition through the reaction with [BMIM]DCA.

Experimental

Note! Chemical experiments require strict safety practices. Reagents must be handled carefully, and their properties must be understood beforehand. For example, in the IL-H₂O₂ ignition tests, although 98% H₂O₂ vapor is not significantly toxic, its strong oxidizing nature demands caution. Direct skin contact can damage cell structures and cause severe burns. When metal complexes are used to catalyze the IL-H₂O₂ reactions, there is also a risk of accidental ignition of flammable materials. H₂O₂ quantities must be strictly controlled, as excess can rapidly produce large gas volumes, potentially leading to explosions. Therefore, comprehensive safety measures must be implemented, including performing all operations in a functioning fume hood, wearing safety goggles, and preventing gas accumulation.

Materials

All chemicals were commercially available analytical grade reagents and were used without further purification. 1-Butyl-3-methylimidazolium dicyandiamide ([BMIM]DCA, 98%) was purchased from Shanghai Chengjie Chemical Co., Ltd. Manganese chloride (MnCl₂, 99%) was bought from Damas-Beta. Ferrous chloride (FeCl₂, 99.5%), copper chloride (CuCl₂, 98%), vinyl imidazole (vim, 99%), sodium dicyandiamide (NaDCA, 96%), and fluoroboric acid (HBF₄, 40%) were from Aladdin Company. Cobalt chloride (CoCl₂, 99.7%) was from Marklin. Triethylamine (TEA, 99%) was from General-Reagent. Hydrogen peroxide (H₂O₂, 98%) was purchased from Shanghai Xushi Science and Technology Co., Ltd.

Synthetic procedures

The synthesis of M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) is illustrated in Fig. 2. The detailed procedures are as follows:

Fig. 2 Synthesis of the M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) complexes.

Mn(vim)₂(DCA)₂ complex. MnCl₂ (2 mmol, 0.252 g) was dissolved in H₂O (10 mL) with magnetic stirring. EtOH (10 mL) was added, and the mixture was heated. At 60 °C, vinylimidazole (4 mmol, 0.376 g) and 1.33 M NaDCA solution were added dropwise, followed by the addition of HBF₄ (200 µL) as an acid regulator. The reaction was kept at 60 °C for 2 h, then filtered to give a colourless solution. Colourless block crystals formed after standing at room temperature for 2 days.

Fe(vim)₂(DCA)₂ and Co(vim)₂(DCA)₂ complexes. These were prepared similarly to the Mn complex. FeCl₂ (2 mmol, 0.254 g) or CoCl₂ (2 mmol, 0.259 g) was dissolved in H₂O (10 mL), followed by the addition of EtOH (10 mL). The mixture was heated to 60 °C with stirring, then vinylimidazole (4 mmol, 0.376 g), 1.33 M NaDCA solution, and HBF₄ (200 µL) were added dropwise. After stirring at 60 °C for 2 h, the mixture was filtered while the solvent was still hot. The filtrate of Fe(vim)₂(DCA)₂ was colorless and that of Co(vim)₂(DCA)₂ was pink. After standing at room temperature for 2 days, colorless rod-like crystals (Fe) and orange block crystals (Co) were obtained.

Cu(vim)₂(DCA)₂ complex. CuCl₂ (2 mmol, 0.269 g) was dissolved in H₂O (10 mL). DMF (10 mL) was added, and the mixture was heated to 60 °C with magnetic stirring. At this temperature, vinylimidazole (4 mmol, 0.376 g) and 1.33 M NaDCA solution were added dropwise, followed by the addition of TEA (150 µL). The reaction was stirred at 60 °C for 2 h. After hot filtration, a clear blue filtrate was obtained. The blue needle-like crystals were formed after standing at room temperature for 2 days.

Characterization

Single crystal X-ray diffraction (SCXRD) data were collected on a Gemini E diffractometer with Mo Kα radiation (λ = 0.71073 Å). Data were processed using the Olex2 software to solve the structure. Powder X-ray diffraction (PXRD) patterns were recorded on a Rigaku/Smart Lab SE diffractometer using Cu Kα₁ and Co Kα₁ radiation, scanning from 5° to 50° at 10° min⁻¹. Phase purity was verified by comparison with the SCXRD-simulated patterns. Mixed fuels were prepared by blending 10 wt% catalyst with 90 wt% [BMIM]DCA for physicochemical tests and ignition evaluation. Thermal stability was assessed on a Mettler TGA3+ thermogravimetric analyzer under N₂ atmosphere, heating from 50 °C to 800 °C at 10 °C min⁻¹. The heat of combustion (ΔH_c) was measured using an oxygen-bomb calorimeter. The long-term dispersion stability of the complexes in IL was studied using a Malvern Panalytical Zetasizer Pro particle-size analyzer and a Shimadzu UV-3600 PLUS spectrophotometer (scan range 240–800 nm). Ignition tests were recorded with a high-speed camera (OLYMPUS i-SPEED3, 1000 fps). Electron paramagnetic resonance (EPR) spectra were acquired on a Bruker EMXPlus-10/12 spectrometer to detect the reactive oxygen species. In addition, the formation enthalpy (ΔH_f) of each complex was computed using the CP2K 2023.1 method (see SI), from which the ΔH_f of the mixed fuel was derived. Specific impulse (I_sp) data for the [BMIM]DCA-H₂O₂ system with 10 wt% catalyst were calculated through the NASA-CEA program with an oxidizer-to-fuel (O/F) range of 0.5–6.

Results and discussion

Catalyst structures

The M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) complexes were synthesized, and the crystallographic data are shown in Tables S1–S5. The complexes were characterized by FT-IR, PXRD and TG. In the FT-IR spectra (Fig. S1), the peak at 2173 cm⁻¹ corresponds to C≡N of the DCA ligand. The peaks at 3137 cm⁻¹, 1358 cm⁻¹, and 1102 cm⁻¹ are assigned to the unsaturated C–H stretching vibration, C–C/C–N skeleton stretching vibration, and C–H bending vibration of the imidazole ring, respectively. The peak at 1647 cm⁻¹ represents the C=C stretching of the vinyl group on the imidazole ring. The PXRD patterns (Fig. S2) show that the experimental diffraction peaks match the simulated patterns, indicating high purity of the sample. TG curves (Fig. S3) show decomposition temperatures (T_d) of 186 °C (Mn), 218 °C (Fe), 218 °C (Co), and 202 °C (Cu), demonstrating their good thermal stabilities.

The SCXRD structures of M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) are shown in Fig. 3. The Mn and Fe complexes crystallize in the triclinic space group P1̄, and the Co complex in the orthorhombic space group Cmca. They exhibit similar coordination patterns. Each M(II) center is coordinated by six N atoms in an octahedral geometry and linked via DCA bridges, forming one-dimensional (1D) chain structures. In Mn(vim)₂(DCA)₂, the Mn(II) ion coordinates with the N atom at the 3rd position of the imidazole ring in vim, and the Mn–N bond lengths are both 2.2 Å. The coordination environment of Mn(II) is the same as that of the N atom of C≡N in DCA, and the Mn–N bond length is 2.3 Å. The N–Mn–N bond angles vary from 87.9° to 180.0°. The C–N–Mn bond angles range from 126.5° to 163.1°. In Fe(vim)₂(DCA)₂, the Fe(II) ion coordinates with the N atom of the imidazole ring in vim, and the bond length is 2.2 Å. The Fe(II) ions alternately coordinate with the N atom of the C≡N group in DCA, and the bond length is 2.2 Å. The N–Fe–N bond angles range from 88.0° to 180.0°. The C–N–Fe bond angles range from 126.2° to 163.3°. In Co(vim)₂(DCA)₂, Co(II) forms a coordination bond with the N atom of the imidazole ring in vim, and the bond length is 2.1 Å. Co(II) forms a coordination bond with the N atom of the C≡N group in DCA, and the bond length is 2.1 Å. The N–Co–N bond angle ranges from 89.2° to 180.0°. The C–N–Co bond angles range from 126.7° to 160.8°.

Fig. 3 Structures of (a) Mn(vim)₂(DCA)₂, (b) Fe(vim)₂(DCA)₂, (c) Co(vim)₂(DCA)₂, and (d) Cu(vim)₂(DCA)₂. Hydrogen atoms have been omitted for clarity.

Cu(vim)₂(DCA)₂, in the triclinic space group P1̄, features a four-coordinate planar quadrilateral geometry, resulting in a zero-dimensional (0D) isolated molecule. Each asymmetric unit contains one M atom and two vim ligands. The N atoms at the symmetric positions of the Cu center are equivalent. There are two sets of bond lengths and bond angles of the same type. The Cu–N bond lengths are both 2.0 Å. The N–Cu–N bond angles of adjacent ligands are 90.2° and 89.8°, respectively, which are in proximity to a right angle. The N–Cu–N bond angles between the same ligands are both 180.0°, confirming a planar geometry. All structures assemble into three-dimensional networks via van der Waals interactions.

Physicochemical properties of the fuels (10 wt% complex + 90 wt% IL)

The M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) complexes were ground into powder form in a mortar. For each complex, 0.02 g was added to [BMIM]DCA (0.18 g), and the mixture was ultrasonicated for 30 minutes to obtain a uniformly dispersed fuel. The properties of these fuels are summarized in Table 1.

Table 1 Physicochemical properties of the fuels (10 wt% catalyst + 90 wt% IL)

Propellant fuels	Td^a (°C)	ΔHc^b (MJ kg^−1)	ΔHf^c (KJ mol^−1)	Isp^d (s)	tid^e (ms)
a Decomposition temperature.b Heat of combustion.c Formation enthalpies.d Specific impulse (O/F = 3.5, F is a mixed fuel and O is 98% H2O2).e Ignition delay time (average value of three tests).
[BMIM]DCA (IL)	288	—	245.0	267.5	No
Mn(vim)2(DCA)2 + IL	251	29.6	193.3	253.0	Bubble
Fe(vim)2(DCA)2 + IL	242	—	195.9	252.9	61
Co(vim)2(DCA)2 + IL	249	28.8	200.1	252.9	Bubble
Cu(vim)2(DCA)2 + IL	267	32.3	204.2	253.0	21

---

After catalyst addition, the Cu-based fuel exhibits the highest T_d of 266.7 °C (Fig. 4a). The heat of combustion (ΔH_c) was measured by oxygen-bomb calorimetry (Fig. S5a), and the values for Mn-, Co-, and Cu-based mixed fuels are 29.6, 28.8, and 32.3 MJ kg⁻¹, respectively. The Fe-based sample could not be burned by O₂ during the test, and no relevant data were obtained. The ΔH_f of each mixed fuel was derived (Fig. 4b) through Table S6 and eqn (1)–(4) (SI), and the Cu-based mixed fuel shows the highest ΔH_f of 204.2 kJ mol⁻¹. Specific impulse (I_sp) was calculated using NASA-CEA with chamber pressure P_c = 300 psia and nozzle expansion ratio P_c/P_e = 20.4. Upon catalyst addition, all mixed fuels exhibit a similar I_sp value of 253.0 s (O/F = 3.5, Fig. S5b), as the complexes share the same ligands.

Fig. 4 (a) Decomposition temperature and (b) heat of formation of the fuels (10 wt% catalyst + 90 wt% IL).

Stability of 10 wt% complex in [BMIM]DCA

After ultrasonication, the mixed fuels are transparent liquids. To determine whether the complexes form colloidal dispersions or true solutions in [BMIM]DCA, Tyndall experiments were performed. As shown in Fig. 5 and Fig. S7, pure [BMIM]DCA shows no Tyndall effect, while all complexes + [BMIM]DCA mixtures exhibit clear Tyndall beams, confirming colloidal dispersion rather than dissolution.

Fig. 5 Tyndall effect of Fe(vim)₂(DCA)₂ + IL and Cu(vim)₂(DCA)₂ + IL.

Long-term dispersion stability was evaluated using dynamic light scattering (DLS) and UV-vis spectroscopy with Cu(vim)₂(DCA)₂ as an example. DLS (Fig. S8) shows that the particle sizes of Cu(vim)₂(DCA)₂ are 369.6 nm in [BMIM]DCA. UV-vis absorption spectra of M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) + [BMIM]DCA mixtures are shown in Fig. 6 and S9. Compared to the solid-state UV-vis spectra of pure complexes (Fig. S4), the absorption bands of mixtures are red-shifted due to the solvation of [BMIM]DCA.^31,32 Strong absorption bands appear at 346 (Mn), 362 (Fe), 288 (Co), and 359 nm (Cu), assigned to the charge-transfer transitions of π → M(II).^33–36 The strong absorption bands were monitored over 25 days (seven measurements) and showed no significant changes in position or intensity, confirming the long-term dispersion stability of complexes in [BMIM]DCA.

Fig. 6 (a) UV-vis spectrum of Fe(vim)₂(DCA)₂ + [BMIM]DCA. (b) Absorbance plot of Fe(vim)₂(DCA)₂ + [BMIM]DCA at 362 nm. (c) UV-vis spectrum of Fe(vim)₂(DCA)₂ + [BMIM]DCA. (d) Absorbance plot of Cu(vim)₂(DCA)₂ + [BMIM]DCA at 359 nm.

Ignition delay time measurements

Pure [BMIM]DCA and H₂O₂ are non-hypergolic. In the three-component system (IL; 98% and H₂O₂; catalyst), the ignition is primarily driven by the oxidizer H₂O₂. It means that the catalyst mainly promotes the decomposition of H₂O₂ into ROS, which then triggers the ignition with ionic liquid. The catalytic ignition performance of M(vim)₂(DCA)₂ in the [BMIM]DCA-H₂O₂ was evaluated by drop tests (Table 1). A droplet of mixed fuel (10 wt% complex + 90 wt% [BMIM]DCA) was dropped into a vial containing 1.5 mL of 98% H₂O₂. Ignition was recorded with a high-speed camera at 1000 fps. The ignition delay time (t_id) is defined as the interval from droplet-liquid contact to the appearance of an open flame, and values are the average of three tests. The results show that only rapid bubbling was observed for Mn- and Co-based fuels. As shown in Fig. 7, the Fe-based fuel has an average t_id of 61 ms, with a standard deviation of ±1.0 ms. In comparison, Cu(vim)₂(DCA)₂ exhibits better catalytic performance with an average t_id of 21 ms and a standard deviation of ±0.5 ms.

Fig. 7 Ignition delay times of (a) 10 wt% Fe(vim)₂(DCA)₂ + IL and (b) 10 wt% Cu(vim)₂(DCA)₂ + IL with H₂O₂ recorded using a high-speed camera (1000 fps).

Comparison of the t_id of IL-H₂O₂ between the reported and our catalysts is shown in Fig. 8. Iodine-containing catalysts, such as 10 wt% [N(CH₃)₄]_n(Cu₂I₃)_n (t_id = 55 ms)¹⁹ and 10 wt% [FcCH₂N(CH₃)₃]₂(Cu₂I₄) (t_id = 54 ms)¹⁹ in [EMIM][BH₃CN]-H₂O₂, were found to have relatively longer ignition delays. In these systems, iodine is believed to lower the Gibbs free energy of the reaction between H₂O₂ and [EMIM][BH₃CN], thereby promoting the reaction.^37,38 Several non-iodine catalysts have been reported with shorter ignition delays, including 10 wt% Cu(AIM)₄(BH₃CN)₂ (t_id = 37 ms),³⁹ 13 wt% Cu₂[C(CH₃)COO]₄·2CH₃OH (t_id = 30 ms),²⁹ and 10 wt% Cu(vim)₂(DCA)₂ (t_id = 21 ms) in [EMIM][BH₃CN]/[BMIM]DCA-H₂O₂. The improved performance in these systems is generally attributed to Cu(II) directly catalyzing H₂O₂ decomposition to generate ROS, which enables rapid ignition with the ILs. In this work, low-dimensional coordination-unsaturated complexes, such as Cu(vim)₂(DCA)₂, offer fully exposed active sites. Concurrently, the vim ligand modulates the electron density of the metal center and facilitates direct H₂O₂ adsorption and activation, ultimately yielding faster ignition.

Fig. 8 Comparison of the t_id of IL-H₂O₂ with reported catalysts.

Analysis of the catalytic mechanism

EPR spectroscopy was applied to monitor the reactive oxygen species (ROS) generated during the catalytic process. For safety, tests were conducted using 30% H₂O₂. As shown in Fig. 9, the ROS (·OH, ·O₂⁻, and ¹O₂) produced upon catalysis by 10 wt% Cu(vim)₂(DCA)₂ + 90 wt% [BMIM]DCA were trapped with 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) for ·OH/·O₂⁻ and 2,2,6,6-tetramethylpiperidine (TEMP) for ¹O₂.^40–42 The lifetime of ·OH is very short (<1 µs) and its diffusion ability is limited, although it possesses the highly oxidizing ability.⁴³ In contrast, ·O₂⁻ and ¹O₂ have millisecond-scale lifetimes and can diffuse over longer distances, enabling sustained and effective reaction with [BMIM]DCA.^44–46 Hence, ·O₂⁻ and ¹O₂ are considered the primary ROS responsible for the ignition of the [BMIM]DCA-H₂O₂ propellant.

Fig. 9 Signals of the ROS (·OH, ·O₂⁻ and ¹O₂) during the catalytic decomposition of [BMIM]DCA-H₂O₂ by Cu(vim)₂(DCA)₂.

To investigate the weak interactions in the Cu(vim)₂(DCA)₂ complex, an Independent Gradient Model (IGM) analysis was performed. Fig. 10a and S11 show the 3D structure of the complex from the top and side views. Green isosurfaces are clearly visible between the imidazole rings, which indicate the π–π stacking interactions. The electron distribution of Cu(vim)₂(DCA)₂ was then calculated (Fig. 10b), as the HOMO is mainly distributed on the ligand and the LUMO is located near the Cu site. In addition, the GAP value of the catalyst is only 0.306, and electron transfer in the catalyst is relatively facile, thereby promoting the catalytic activity at the Cu site. The density of states (DOS) calculations show that the Fermi level (E_F) is embedded into the valence band maximum (VBM), giving the complex metallic properties to some degree (Fig. 10c). This feature enables efficient electron transfer through π–π conjugation for the efficient activation of H₂O₂ during the catalytic process.

Fig. 10 (a) IGM isosurfaces of Cu(vim)₂(DCA)₂. (b) DFT-calculated HOMO and LUMO of Cu(vim)₂(DCA)₂. (c) Density of states of Cu(vim)₂(DCA)₂.

The catalytic activation of H₂O₂ by Cu(vim)₂(DCA)₂ is illustrated through a proposed cycle in Fig. 11a. The planar quadrilateral structure of Cu(vim)₂(DCA)₂ exposes the axial (z-direction) active sites of Cu(II). The process begins with H₂O₂ adsorption onto the Cu(II) center along the z-axis to form intermediate I. Subsequently, weak coordination of the second oxygen lone pair to Cu(II) yields intermediate II, followed by cleavage of the O–O bond to generate intermediate III containing two Cu–O bonds. One Cu–O bond then breaks, releasing a hydroxyl radical (·OH) and producing intermediate IV. The ·OH group participates in hydrogen bonding with another H₂O₂ molecule to form intermediate V. Subsequently, electron transfer within the H–O bond of H₂O₂ generates a proton, ·O₂⁻ and ¹O₂, thus forming intermediate VI. Cu(II)–H₂O releases a water molecule, and the catalyst returns to its original Cu(vim)₂(DCA)₂ structure.²⁹ The generated ROS (·OH, ·O₂⁻, and ¹O₂) attack the [BMIM]⁺ and DCA⁻ ions of IL, and initiate chain-oxidation decomposition to CO₂, H₂O, N₂, thereby achieving rapid ignition.

Fig. 11 (a) Proposed catalytic ignition process of Cu(vim)₂(DCA)₂ with H₂O₂. (b) Gibbs free energy changes in Cu(vim)₂(DCA)₂.

Gibbs free energy of the catalytic ignition process was systematically studied by the DFT method (Fig. 11b). The results show that H₂O₂ was adsorbed onto the Cu site with an adsorption energy of −0.38 eV, forming hydrogen bonds with the N atoms of the ligand. The subsequent O–O bond cleavages to form 2 *OH with an energy barrier of 0.52 eV. After releasing one ·OH radical, a second H₂O₂ molecule interacts with the remaining *OH via hydrogen bonding. The hydrogen transfer from H₂O₂ to *OH occurs with a barrier of 0.40 eV and generates *H₂O and *OOH. Finally, the desorption of H₂O regenerates the active catalytic sites. The unsaturated Cu center and electron-rich ligands improve the production of reactive oxygen species, resulting in the rapid ignition of the [BMIM]DCA-H₂O₂ propellant.

Conclusions

A series of complexes, M(vim)₂(DCA)₂ (M = Mn, Fe, Co, and Cu) catalysts, were synthesized. X-ray single crystal diffraction shows that the Mn, Fe, and Co complexes adopt 1D octahedral structures, while Cu(vim)₂(DCA)₂ is a 0D planar quadrilateral complex. The physicochemical properties of mixed fuels consisting of 10 wt% M(vim)₂(DCA)₂ and 90 wt% [BMIM]DCA were measured or calculated. The Cu-based mixed fuel exhibits comprehensive performance: T_d = 266.7 °C, ΔH_c = 32.3 MJ kg⁻¹, and ΔH_f = 204.2 KJ mol⁻¹. After the addition of the catalyst, all fuels exhibited similar specific impulse (I_sp) values (253 s, O/F = 3.5). Dispersion stability tests indicated that Cu(vim)₂(DCA)₂ forms a colloidal dispersion in [BMIM]DCA, and UV-vis spectroscopy confirmed the long-term stable dispersion of all complexes without sedimentation over 25 days. In catalytic ignition tests, Cu(vim)₂(DCA)₂ gave the shortest ignition time with t_id = 21 ms. EPR studies revealed that Cu(vim)₂(DCA)₂ catalyzes H₂O₂ to generate ·OH, ·O₂⁻, and ¹O₂, which subsequently trigger the rapid chemical ignition of [BMIM]DCA. This high activity was further investigated by DFT calculations. The zero-dimensional geometry of Cu(vim)₂(DCA)₂, with its unsaturated Cu center and electron-rich ligand framework, enables efficient electron transfer and promotes the generation of reactive oxygen species, leading to fast ignition. This work elucidates the pivotal role of the metal center in the IL-H₂O₂ system and offers valuable insights into the catalytic ignition of green propellants.

Author contributions

Yuqi Zhang: conceptualization, data curation, formal analysis, investigation, methodology, validation, visualization, and writing original draft. Yuan Yao: software, formal analysis, and visualization. Qing Cheng: supervision, resources, and writing – review and editing. Yanqiang Zhang: funding acquisition, project administration, resources, and writing – review and editing.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

CCDC 2517002 {Fe(vim)₂(DCA)₂}, 2517003 {Mn(vim)₂(DCA)₂}, 2517004 {Co(vim)₂(DCA)₂} and 2517005 {Cu(vim)₂(DCA)₂} contain the supplementary crystallographic data for this paper.^47a–d The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: materials, computational methods, Fig. S1–S11, and Tables S1–S6. See DOI: https://doi.org/10.1039/d6ra00294c.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (U21A20307 and 22178359) and the CAS Project for Young Scientists in Basic Research (YSBR-052).

Notes and references

1. F. S. Forbes, and P. A. Van Splinter, Encyclopedia of Physical Science and Technology, 2003, vol. 3, pp. 741–777.
2. S. M. Davis, and N. Yilmaz, Advances in Aerospace Engineering, 2014, p. 729313.
3. W. C. Keller, Aviat., Space Environ. Med., 1988, 59, A100–A106.
4. E. H. Vernot, J. D. Macewen, R. H. Bruner, C. C. Haun, E. R. Kinkead, D. E. Prentice, A. Hall, R. E. Schmidt, R. L. Eason, G. B. Hubbard and J. T. Young, Fundam. Appl. Toxicol., 1985, 5, 1050–1064.
5. W. Daimon, M. Tanaka and I. Kimura, Symp. (Int.) Combust., [Proc.], 1985, 20, 2065–2071.
6. A. Wilson, 2nd International Conference on Green Propellants for Space Propulsion, ESA Special Publication, 2004, p. 557.
7. S. Schneider, T. Hawkins, M. Rosander, G. Vaghjiani, S. Chambreau and G. Drake, Energy Fuels, 2008, 22, 2871–2872.
8. Q. Zhang and J. M. Shreeve, Chem. Rev., 2014, 114, 10527–10574.
9. I. I. Sam, S. Gayathri, G. Santhosh, J. Cyriac and S. Reshmi, J. Mol. Liq., 2022, 350, 118217.
10. R. Ciriminna, L. Albanese, F. Meneguzzo and M. Pagliaro, Chemsuschem, 2016, 9, 3374–3381.
11. V. K. Bhosale, K. Lee, H. Yoon and S. Kwon, Fuel, 2024, 376, 132688.
12. S. D. Rosenberg, Acta Astronaut., 1982, 9, 663–669.
13. K. R. Gagne, M. R. McDevitt and D. L. Hitt, Aerospace, 2018, 5, 52.
14. J. L. Shamshina, M. Smiglak, D. M. Drab, T. G. Parker, H. W. H. Dykes, R. D. Salvo, A. J. Reich and R. D. Rogers, Chem. Commun., 2010, 46, 8965–8967.
15. Y. T. Zhang, X. Zhang, D. Y. Yin and Q. H. Zhang, Molecules, 2025, 30, 1789.
16. V. K. Bhosale, J. Jeong, J. Choi, D. G. Churchill, Y. Lee and S. Kwon, Combust. Flame, 2020, 214, 426–436.
17. Q. Cheng, Y. Q. Zhang, Y. Y. Cao, P. H. Dou, L. Liu and Y. Q. Zhang, J. Mater. Chem. A, 2025, 13, 17434–17442.
18. K. L. Xie, X. Zhuang, X. Luo, Z. Y. Jing, X. Y. Song, A. Q. Hou and A. Q. Gao, Green Chem., 2023, 25, 4438–4445.
19. K. C. Wang, T. L. Liu, Y. H. Jin, S. Huang, N. Petrutik, D. Shem-Tov, Q. L. Yan, M. Gozin and Q. H. Zhang, J. Mater. Chem. A, 2020, 8, 14661–14670.
20. Z. Li, X. Q. Qi, J. J. Wang, Z. Z. Zhu, J. X. Jiang, X. B. Niu, A. Cabot, J. S. Chen and R. Wu, Susmat, 2023, 3, 498–509.
21. W. Liu, X. Q. Ding, J. J. Cheng, J. L. Jing, T. S. Li, X. Huang, P. P. Xie, X. C. Lin, H. L. Ding, Y. Kuang, D. J. Zhou and X. M. Sun, Angew. Chem., 2024, 136, e202406082.
22. P. Deshlahra, R. T. Carr and E. Iglesia, J. Am. Chem. Soc., 2014, 136, 15229–15247.
23. J. Ha, M. Jeon, J. Park, J. Kim and H. R. Moon, Nanoscale Adv., 2023, 5, 2111–2117.
24. C. Xu, M. He, B. B. Chen and B. Hu, ACS Appl. Nano Mater., 2023, 6, 17521–17530.
25. K. F. Yu, P. He, D. Zhang, H. Q. Wu, K. W. Shu, H. H. Long, Y. B. Sun, X. Pei, B. Jiang, K. C. Cao, Y. Li and L. J. Ma, Angew. Chem., Int. Ed., 2025, 64, e202511501.
26. P. Wu, F. F. Fang, J. S. Song, W. Peng, J. Liu, C. S. Li, Z. X. Cao and B. J. Wang, J. Am. Chem. Soc., 2019, 141, 19776–19789.
27. X. P. Zhang, A. Chandre, Y. M. Lee, R. Cao, K. Ray and W. Nam, Chem. Soc. Rev., 2021, 50, 4804–4811.
28. S. M. Adam, I. Garcia-Bosch, A. W. Schaefer, S. K. Sharma, M. A. Siegler, E. I. Solomon and K. D. Karlin, J. Am. Chem. Soc., 2017, 139, 472–481.
29. Q. Cheng, Y. Yao, P. H. Dou, L. Liu, Y. Y. Cao and Y. Q. Zhang, J. Mater. Chem. A, 2024, 12, 21425–21433.
30. F. He, Y. N. Liu, X. Y. Peng, Y. Q. Chen, Q. Zheng, B. Yang, Z. J. Li, Q. Zhou, Q. H. Zhang, J. G. Lu, L. C. Lei, G. Wu and Y. Hou, Nat. Commun., 2025, 16, 6974.
31. C. H. Wang, W. Zhou, J. M. Ma, Z. Wang and C. Y. Zhang, Molecules, 2025, 30, 2352.
32. J. Zhang, J. Yao and H. R. Li, J. Phys. Chem. B, 2022, 126, 2279–2284.
33. C. A. Tibbetts, A. B. Wyatt, B. M. Luther, A. K. Rappé and A. T. Krummel, J. Phys. Chem. B, 2023, 127, 932–943.
34. A. B. P. Lever, J. Chem. Educ., 1974, 51, 612–616.
35. M. Gras, N. Papaiconomou, E. Chainet and I. Billard, Solvent Extr. Ion Exch., 2018, 36, 583–601.
36. S. P. Devi, N. S. Devi, L. J. Singh, R. K. B. Devi, W. R. Devi, C. B. Singh and R. K. H. Singh, Inorg. Nano-Met. Chem., 2017, 47, 223–233.
37. F. A. S. Mota, L. H. Fei, C. L. Tang, Z. H. Huang and F. S. Costa, Fuel, 2022, 336, 127086.
38. J. J. Pei, J. B. Liu, K. X. Fu, Y. K. Fu, K. Yin, S. L. Luo, D. Y. Yu, M. Y. Xing and J. M. Luo, Nat. Commun., 2025, 16, 800.
39. X. Zhao, Z. Wang, X. J. Qi, S. W. Song, S. Huang, K. C. Wang and Q. H. Zhang, Inorg. Chem., 2021, 60, 17033–17039.
40. A. J. Carmichael, L. Steelgoodwin, B. Gray and C. M. Arroyo, Free Radical Res. Commun., 1993, 19, S1–S16.
41. K. Reszka, P. Bilski and C. F. Chignell, Free Radical Res. Commun., 1992, 17, 377–385.
42. J. H. Wu, F. Chen, T. H. Yang and H. Q. Yu, Proc. Natl. Acad. Sci. U. S. A., 2023, 120, e2305706120.
43. J. W. E. Wilson, J. Phys. Chem. Ref. Data, 1972, 1, 535–573.
44. M. G. Campo and J. R. Grigera, J. Chem. Phys., 2005, 123, 084507.
45. H. R. Yu, D. X. Liu, H. Y. Wang, H. S. Yu, Q. Y. Yan, J. H. Ji, J. L. Zhang and M. Y. Xing, Chin. J. Catal., 2022, 43, 2678–2689.
46. M. M. Islam, T. Imase, T. Okajima, M. Takahashi, Y. Niikura, N. Kawashima, Y. Nakamura and T. Ohsaka, J. Phys. Chem. A, 2009, 113, 912–916.
47. (a) CCDC 2517002: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2qh4m0; (b) CCDC 2517003: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2qh4n1; (c) CCDC 2517004: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2qh4p2; (d) CCDC 2517005: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2qh4q3.

---