Synthesis of Thermal-Stable Aviation Fuel Additives with 4-Hydroxy-2-butanone and Cycloketones

Abstract

A novel two-step strategy was developed for the efficient synthesis of decalin and octahydroindene from lignocellulose-derived platform compounds. In the first step, bicyclic intermediates were directly generated via a cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone with cyclohexanone (or cyclopentanone). Among the evaluated catalysts, CaO demonstrated the highest activity and selectivity. Based on CO₂-TPD results, the excellent performance of CaO can be rationalized by its proper basicity. In the second step, these intermediates were selectively hydrodeoxygenated to decalin (or octahydroindene) over Ni/H-ZSM-5 catalyst. Under the investigated reaction conditions, ~90% overall yields of decalin and octahydroindene were achieved. This work provides a viable strategy for the selective conversion of lignocellulose-derived platform compounds to the additives for improving the thermal stability of aviation fuel.

1. Introduction

Due to the increasing of social concerns about energy and environmental issues, the synthesis of jet fuel range hydrocarbons with renewable, cheap and, abundant lignocellulose has drawn a lot of attention [1,2,3,4,5,6,7,8,9,10,11,12]. Decalin and octahydroindene are two major components of jet propellant-900 (JP-900), a coal-based thermal-stable aviation fuel [13]. At the same time, these compounds are also widely used as the additives to improve the thermal-stability and volumetric heat value of aviation fuels [13,14,15] or the intermediates in the manufacture of many chemicals, medicines, or shockproof agent in rubber industry. During the past years, great efforts have been devoted to the synthesis of renewable decalin and octahydroindene with the platform compounds that can be derived from lignocellulose [16,17,18,19].

4-Hydroxy-2-butanone is a platform compound that can be synthesized by the aldol condensation of formaldehyde and acetone from the acetone-butanol-ethanol (ABE) fermentation of lignocellulose [20,21]. Cyclohexanone and cyclopentanone are two cycloketones that can be obtained from the aqueous-phase selective hydrogenation of furfural [22,23,24,25,26,27,28] or the selective hydrogenation of phenol [29,30,31,32,33] from the hydrogenolysis of lignin [34,35]. In the recent work of our group [36,37], it was found that cyclopentanone can also be obtained by the direct hydrogenolysis of xylose and hemicellulose in a NaCl aqueous solution/toluene biphasic system. In the previous work of our group [17,38], decalin was synthesized by the aldol-condensation of cyclopentanone, followed by hydrogenation/rearrangement and hydrogenation (or hydrodeoxygenation and isomerization). Cyclopentanol is the by-product generated during the production of bio-cyclopentanone. As another option, decalin and its alkylated analogs were also prepared by Zou et al. [18] and our group [16] by the alkylation/rearrangement of cyclopentanol with alkylated cycloalkanes (or the dehydration, dimerization/rearrangement, and hydrogenation of cyclopentanol). In this work, decalin and octahydroindene were selectively synthesized by the cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclohexanone and cyclopentanone, followed by hydrodeoxygenation. The strategy of this work was illustrated in Scheme 1. In the first step, cyclohexanone and 4-hydroxy-2-butanone reacted at 403 K for 6 h in the presence of CaO catalyst at an initial 4-hydroxy-2-butanone/cyclohexanone molar ratio of 1:4, affording 4,4a,5,6,7,8-hexahydronaphthalen-2(3H)-one (HHN) with complete conversion of 4-hydroxy-2-butanone and a HHN molar yield of 96.8%. Subsequently, hydrodeoxygenation of HHN over a 10%Ni/H-ZSM-5 catalyst produced decalin with a molar yield of 91.7%. Under the same conditions, the reaction of cyclopentanone and 4-hydroxy-2-butanone afforded 1,2,3,6,7,7a-hexahydro-5H-inden-5-one (THI) in 98.3% yield. This compound was subsequently hydrodeoxygenated to octahydroindene with a molar yield of 89.6%.

2. Results and Discussion

As the first innovation of this work, 4,4a,5,6,7,8-hexahydronaphthalen-2(3H)-one (HHN), a C₁₀ oxygenate, was obtained by the reaction of 4-hydroxy-2-butanone and cyclohexanone over a series of alkaline earth metal oxide or magnesium aluminum hydrotalcite (MgAl-HT) catalysts (see Figure 1 and Figures S1–S6). From Scheme 1, we can see that the HHN obtained has the same carbon chain structure as that of decalin. Therefore, it can be potentially used as a precursor for the lignocellulose-based renewable decalin. Among the investigated solid base catalysts, CaO exhibited the best performance (see Figure 1 and Figure S1). Over it, good cyclohexanone conversion 40.0% and higher HHN yield 38.3% were achieved after the reaction was carried out at 383 K for 4 h. It is worth mentioning that 2-cyclohexylidenecyclohexanone (generated from the self-aldol condensation of cyclohexanone) was also obtained as the by-product over the SrO and BaO catalysts (see Figures S7 and S8). This may be the reason why lower HHN yields were achieved over these catalysts although the cyclohexanone conversions over the SrO and BaO catalysts were even higher than that over the CaO catalyst.

To gain deeper insight into the excellent performance of CaO, we characterized the investigated catalysts by N₂-physisorption, CO₂-chemisorption, and CO₂-TPD. From the results illustrated in

Table 1. Specific BET surface areas (S_BET) and the amounts of base sites of the investigated catalysts.

Catalyst	SBET (m2 g−1) 1	Amount of Base Sites (μmol g−1) 2
MgO	44.4	4.8
MgAl-HT	202.0	6.5
CaO	8.9	13.7
SrO	5.1	5.3
BaO	5.6	74.8

¹ Measured by N₂-physisorption; ² Measured by CO₂-chemisorption.

, no evident relationship was observed between the specific BET surface areas or the amounts of base sites of the investigated solid base catalysts and their performances for the condensation of 4-hydroxy-2-butanone and cyclohexanone. In contrast, an evident corresponding relationship was noticed between the cyclohexanone conversions over the investigated solid base catalysts and their base strengths that were indicated by the desorption temperatures of CO₂ from the surfaces of these catalysts. The CO₂-TPD technique is a crucial tool for assessing the strength of basic sites in solid base catalysts. In general, higher CO₂ desorption temperatures correspond to stronger basic sites on the catalyst surface. As shown in Figure 2, MgO and MgAl-HT exhibit CO₂ desorption peaks in the temperature range of 573–773 K, corresponding to medium-strength basic sites. BaO, SrO, and CaO display CO₂ desorption peaks at higher temperatures (873–1173 K), indicating the presence of strong basic sites on the surfaces of these catalysts. Based on this phenomenon, we can see that higher base strength is favorable for the activation of cyclohexanone. However, a too-high base strength will lead to the self-aldol condensation of cyclohexanone that will compete with the condensation of 4-hydroxy-2-butanone and cyclohexanone. As a result, the CaO catalyst with the moderate base strength demonstrated the best performance for the condensation of 4-hydroxy-2-butanone and cyclohexanone. Taking into consideration the low cost, wide availability, and good performance of CaO, we think that it is a promising catalyst in future applications.

To understand the reaction mechanism for the condensation of 4-hydroxy-2-butanone and cyclohexanone, we did some additional experiments. Firstly, we decreased the reaction temperature. From the analysis of the product obtained at 353 K, 2-(3-oxobutyl)cyclohexanone was identified as the intermediate (see Figures S9 and S10). Subsequently, we studied the reaction of 4-hydroxy-2-butanone over the CaO catalyst in the absence of cyclohexanone. Under the investigated conditions, but-3-en-2-one from the dehydration of 4-hydroxy-2-butanone was identified as the product (see Figures S11 and S12). Based on the experimental results and literature [39,40], a reaction mechanism was proposed in Scheme 2. As the first step, 4-hydroxy-2-butanone undergoes dehydration to form but-3-en-2-one by an E1cb mechanism. The strong electron-withdrawing effect of the carbonyl group facilitates abstraction of the α-hydrogen by a base, generating a stabilized enolate anion intermediate. This intermediate subsequently eliminates the β-hydroxyl group, producing the stable α,β-unsaturated ketone, but-3-en-2-one. Subsequently, but-3-en-2-one participates in a Robinson annulation with cyclohexanone, yielding the key intermediate 2-(3-oxobutyl)cyclohexanone. Finally, intramolecular aldol condensation of 2-(3-oxobutyl)cyclohexanone produces HHN. In the previous work of Olsen et al. [39] and Okano et al. [40], magnesium chloride/triethylamine and lanthanoid triisopropoxides were used as the catalysts for the Robinson annulation and Aldol condensation of unsaturated ketones, respectively. Compared with these catalysts, the CaO catalyst used in present system is superior, in view of the cost and handling of the catalyst. The higher base strength is favorable for the deprotonation of cyclohexanone and 2-(3-oxobutyl)cyclohexanone, which explains the necessity of strong base sites for the one-pot synthesis of HHN from 4-hydroxy-2-butanone and cyclohexanone.

The impact of the molar ratio between 4-hydroxy-2-butanone and cyclohexanone on HHN yield was investigated. When the ratio was adjusted to 1:4, the HHN yield reached 60.0% (Figure 3). This phenomenon might be rationalized by reaction equilibrium. As we know, the Michael addition step in the Robinson annulation is reversible. Therefore, an excess of cyclohexanone may shift the equilibrium toward the formation of 2-(3-oxobutyl)cyclohexanone. It is worth mentioning that HHN was still obtained as the predominate product. No evident peak of 2-cyclohexylidenecyclohexanone was observed in the GC chromatogram of the product obtained even at the initial 4-hydroxy-2-butanone and cyclohexanone molar ratio of 1:4 (see Figure S13), which is advantageous in real application. The effects of reaction temperature and reaction time on the catalytic performance of CaO were examined as well (Figure 4 and Figure 5). It was noticed that the HHN yield over the CaO catalyst initially increased with the reaction temperature (or reaction time), reached the maximum 96.8% when the reaction was carried out at 403 K for 6 h with an initial 4-hydroxy-2-butanone to cyclohexanone molar ratio of 1:4, then decreased with the further increment of the reaction temperature (or reaction time). These results can be comprehended by the further polymerization of HHN that is unexpected in our work.

To assess the practical viability of CaO, we also checked the reusability of the CaO catalyst. After each usage, the CaO was washed with methanol five times, dried it at 373 K for 1 h, and used again for the next batch of reaction. As shown in Figure 6, the CaO catalyst exhibited good stability under the tested conditions. No significant deactivation was observed over five consecutive reaction cycles.

Analogously, we also studied the applicability of the CaO catalyst for the cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclopentanone, another cycloketone that can be derived from hemicellulose [22,23,24,25,26,27,28,36,37]. As we expected, 98.3% yield of THI was achieved over the CaO catalyst under the optimized reaction condition for the cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclohexanone (see Figure 7 and Figures S14–S18). As a potential application, the THI can be used as a precursor for the synthesis of octahydroindene.

As the final objective of this work, we also explored the synthesis of decalin and octahydroindene by the hydrodeoxygenation (HDO) of HHN and HHI. In our previous work [41], Ni/H-ZSM-5 has been found to be an efficient catalyst for the HDO of biomass derived oxygenates. After the reaction was carried out at 473 K and 6 MPa H₂ for 12 h (the reaction conditions were chosen based on our previous work [41] with some minor modification), high molar yields of decalin (91.7%) and octahydroindene (89.6%) were achieved over the 10wt.% Ni/H-ZSM-5 catalyst, respectively (see Figure 8 and Figures S19–S22). According to literature [42,43,44,45], these compounds have high densities (0.90 g mL⁻¹ and 0.91 g mL⁻¹) and low freezing points (230 K and 220 K). As a potential application, they can be used as high-thermal stable advanced jet fuels or additives to improve the thermal-stability of current jet fuels.

3. Materials and Methods

3.1. Preparation of Catalysts

The MgO, CaO, and BaO catalysts were purchased from Aladdin Ltd. (Aladdin Reagent Co., Ltd., Shanghai, China). The magnesium aluminum hydrotalcite (with a Mg/Al atomic ratio of 2:1, denoted as MgAl-HT) was bought as a commercial product from Macklin agent Ltd. (Macklin Biochemical Co., Ltd., Shanghai, China). The SrO catalyst was supplied by Bied Pharma Ltd. (Bied Pharma Ltd., Shanghai, China). The Ni/H-ZSM-5 catalyst was prepared using the impregnation method. The detail preparation procedure, specific BET surface area (S_BET) and metal dispersion of Ni/H-ZSM-5 (see Table S1) were supplied in Supplementary Materials.

3.2. Activity Tests

The cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclohexanone (or cyclopentanone) were conducted in a batch reactor heated by an oil bath under magnetic stirring. After reaction completion, the system was cooled to ambient temperature. The reaction products were mixed with tridecane as the internal standard, diluted, and filtered before being quantitatively analyzed using a gas chromatograph (Agilent Technologies, Wilmington, DE, USA). equipped with a flame ionization detector (FID) and an HP-5 capillary column (30 m × 0.32 mm, 0.25 μm film thickness).

The HDO reaction was conducted in a 50 mL stainless steel autoclave reactor equipped with magnetic stirring. A typical reaction mixture consisted of 0.1 g HHN (or THI), 0.02 g 10% Ni/H-ZSM-5 catalyst, and 5 mL cyclohexane. Prior to the reaction, the reactor was purged with nitrogen three times to ensure the complete removal of air. The system was then pressurized with hydrogen to 6 MPa. The reaction proceeded at 473 K for 12 h, after which the reactor was rapidly cooled to room temperature using a cold-water bath. The reaction products were mixed with tridecane as the internal standard, diluted, and filtered before being quantitatively analyzed using a gas chromatograph (Agilent Technologies, Wilmington, DE, USA) equipped with a flame ionization detector (FID) and an HP-5 capillary column (30 m × 0.32 mm, 0.25 μm film thickness).

Substrate conversion and product yields were calculated using the following Equations:

$$\begin{array}{c}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{u}\mathrm{b}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\left(\%\right)\\ ={\displaystyle \frac{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{u}\mathrm{b}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e} \mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{e}\mathrm{d}}{\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l} \mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{u}\mathrm{b}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}}}\times 100\%\end{array}$$

$$\begin{array}{c}\mathrm{Y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\left(\%\right)\\ ={\displaystyle \frac{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t} \mathrm{f}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{d}}{\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l} \mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}}\times 100\%\end{array}$$

3.3. Characterization

The specific surface areas of catalysts were characterized by N₂-physisorption measurements using the Brunauer–Emmett–Teller (BET) method. The tests were carried out by a Micromeritics ASAP 2460 instrument (Micromeritics Instrument Corporation, Norcross, GA, USA). Before conducting N₂ physisorption analyses, all catalyst samples underwent vacuum pretreatment at 573 K for 6 h to eliminate surface-adsorbed contaminants and gaseous species.

The basicities of the catalysts were characterized by CO₂ chemisorption and temperature-programmed desorption (CO₂-TPD) using a Micromeritics AutoChem II 2920 analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA). For each measurement, 0.1 g catalyst sample was employed. The sample was pretreated under He flow (423 K) to desorb the moisture and impurities before measurement. After the stabilization of baseline, 5% CO₂-He pulses were introduced until the sample reached saturation adsorption. Subsequently, the catalyst was purged by He flow for 0.5 h to remove physically adsorbed CO₂. The adsorption signal was monitored using a thermal conductivity detector (TCD), and the total CO₂ adsorption capacity was quantified by its consumption. After the saturation adsorption of CO₂, the TPD experiment was conducted in the temperature range from 353 to 1173 K at a rate of 10 K min⁻¹ while maintaining helium carrier flow. CO₂ evolution profiles were recorded by monitoring the m/z 44 channel of the OminiStar mass spectrometer.

Hydrogen chemisorption of the Ni/H-ZSM-5 catalysts were performed with a Micromeritics AutoChem II 2920 Characterization System (Micromeritics Instrument Corporation, Norcross, GA, USA). Before each test, the sample was reduced to 10% H₂/Ar flow at 723 K for 1 h, purged with Ar flow at 783 K for 0.5 h and cooled down in Ar flow to 323 K. After the stabilization of baseline, the H₂ adsorption was carried out by the pulse adsorption of 10% H₂/Ar at 333 K.

4. Conclusions

In summary, a novel strategy was developed for the selective synthesis of decalin and octahydroindene by the cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclohexanone (or cyclopentanone) that can be derived from lignocellulose, followed by hydrodeoxygenation. Among the investigated catalysts, CaO exhibited the best performance and good reusability for the cascade dehydration/Robinson annulation of 4-hydroxy-2-butanone and cyclohexanone, which can be rationalized by its moderate base strength. This catalyst is also applicable for 4-hydroxy-2-butanone and cyclopentanone. Under the optimized reaction conditions, high yields of HHN (96.8%) and THI (98.3%) were achieved over the CaO catalyst. After being hydrodeoxygenated over the 10wt.% Ni/H-ZSM-5 catalyst, high carbon yields of decalin (91.7%) and octahydroindene (89.6%) were achieved, respectively. These compounds have high densities and low freezing points. Therefore, they can be potentially used as high-thermal stable advanced jet fuels or additives to improve the thermal-stability of current jet fuels.