Nitric acid oxidation of kerogen in atmospheric and pressurized systems – basis for the Kerox process

1 Introduction

Kerogen, the insoluble macromolecular organic matter found in sedimentary rocks is the most abundant organic carbon resource on the planet (Dyni, 2003; Vandenbroucke and Largeau, 2007). Despite its wide availability, kerogen has been utilized mainly as an energy resource, from direct combustion to shale oil production and gas extraction, while its potential as a source of chemical feedstocks has been largely overlooked (Liu K. et al., 2025). With EU’s increasing need for material supply security and shorter value chains to enhance economic resilience for industry and in policy, research into valorization pathways for local resources that maximize added value and improve economic and resource efficiency is crucial (European Parliament and The Council of the European, 2024). Accordingly, the present work focuses on the direct valorization of kerogen, aiming to identify conditions that enable its efficient conversion into value-added chemicals. In this context, the Estonian Research and Development, Innovation and Entrepreneurship Strategy 2021–2035 identified the valorization of local resources, such as oil shale kerogen into chemicals, as one of its focus areas (Ministry of Education and Research, Ministry of Economic Affairs and Communications, Republic of Estonia, 2021).

The potential utilization methods for different kerogens are governed by their macromolecular structure, which depend on both its origin (lacustrine, marine, or terrestrial) and thermal maturity (Durand, 1980). Chemical modifications of kerogen, such as oxidation, have previously been studied primarily to elucidate kerogen structure. Figure 1 shows the Lille-Blokker structural model of kukersite kerogen structure with oxidation-labile groups (Mets et al., 2023a).

Historical studies on kukersite’s kerogen oxidation with several oxidants including potassium permanganate (KMnO₄), ruthenium tetroxide (RuO₄), nitric acid (HNO₃) and air oxidation, show that oxidation targets oxidation-labile groups, mainly resorcinol subunits, producing predominantly aliphatic dicarboxylic acids (DCAs) with chain lengths typically ranging from C₄ to C₁₀, as well as mono- and tri-carboxylic acids, as illustrated in Figure 2 (Kogerman, 1931; Fomina and Pobul, 1953a; Fomina and Pobul, 1953b; Fomina and Pobul, 1955; Proskurjakov and Soloveichik, 1961; Bajc et al., 2001; Blokker et al., 2001; Lille et al., 2002). The kinetic studies of wet air oxidation (WAO) of kerogen, suggest a two-stage reaction model, consisting of the fast dissolution of the kerogen and the subsequent slower oxidation of the solubilized organic material to low-molecular-weight carboxylic acids (Mets et al., 2023b). Similar oxidative methodologies have also been applied to kerogens from various oil shales beyond Estonian kukersite, demonstrating that kerogen structure governs both the yield and composition of oxidation products (Djuricic et al., 1971; Vitorović et al., 1980; Vitorović et al., 1988; Boucher et al., 1991; Bajc et al., 2008; Khaddor et al., 2008; Barakat et al., 2012; Wang et al., 2017; 2019; Liu Y. J. et al., 2025). This highlights the potential of oxidative pathways for chemical valorization of kerogen.

Industrial research in Estonia during the 1960s–1980s demonstrated that nitric acid oxidation of kerogen could yield up to 26% DCAs after vacuum-distillation in pilot-scale operation, producing humic-acid like plant growth stimulator as a solid by-product from the distillation step. However, the process had significant drawbacks. It required extensive beneficiation to obtain a 90% kerogen concentrate through multiple hydrocycloning and flotation stages, and the stainless-steel reactors experienced severe corrosion. (Palvadre and Ahelik, 1992). Despite these challenges, the research was discontinued primarily due to political factors rather than technical limitations (Veski and Veski, 2019).

Most recently, the authors investigated wet air oxidation (WAO), as well as a two-step process combining WAO with subsequent nitric acid oxidation. WAO alone produces DCA yields of up to 12%, but the isomerically rich mixtures of DCAs, complicate downstream separation (Kaldas et al., 2020; Kaldas et al., 2021). The best results were achieved through two-step oxidation including an initial WAO step under alkaline conditions to achieve carbon conversions of 30%–40%, followed by nitric acid oxidation of the entire crude mixture at 140 °C for 30 min with 8% HNO₃ solution to yield 23% DCAs (Kaldas et al., 2025). The main drawback of this approach is the added complexity of the two-step process. A summary of the key process parameters and findings reported in previous studies and comparison to the present work can be found in Supplementary Table S1. The present study identifies the key parameters governing nitric acid oxidation and establishes the experimental basis for the development of the Kerox process by re-evaluating historical datasets using modern analytical techniques. The design, operation and performance of the continuous-flow reactor without mechanical agitation are beyond the scope of the present work and will be presented in a separate publication.

Nitric acid oxidation can proceed through several competing reaction pathways, with the dominant mechanism determined by reaction severity and the reaction medium. Nitric acid treatment has been applied to coal desulfurization by oxidizing pyritic and organic sulfur species and partially solubilizing organic matter with increasing temperature promoting solubilization (Alvarez et al., 2003). Under relatively mild conditions (room temperature, 2 days, 15–30 wt% HNO₃), nitric acid treatment of peat and coal, results in the formation of nitrohumic acids in the solid phase, which can be utilized as soil amendments for agricultural applications (Thorn and Cox, 2016; Fatima et al., 2021). In contrast, harsher conditions (high temperature, long reaction times, excess HNO₃) are known to favor oxidation and carbon–carbon bond cleavage with the formation of low-molecular weight oxygenated products (Ogata, 1978). Under these conditions, reactive species, such as nitrous acid (HNO₂) and nitrogen oxides (NO_x), may trigger radical-mediated reactions that can become autocatalytic in the presence of oxidation-labile functional groups or mineral components (Ando et al., 2021). Although nitrated and nitrosated products may occur as intermediates of kerogen oxidation, elevated temperatures are expected to promote their further oxidation rather than accumulation as stable products. Therefore, under the reaction conditions employed in the present study, oxidative degradation leading to low-molecular-weight carboxylic acids is expected to be the dominant reaction pathway.

In this work, a systematic evaluation of key parameters affecting nitric acid oxidation of kerogen to yield dicarboxylic acids were evaluated in atmospheric and pressure reactors. The use of nitric acid as an oxidant demands careful optimization to balance high oxidative efficiency against the formation of residual products. This work provides a basis for the development of oxidative pathways for the valorization of kukersite kerogen into higher-value chemical products, to improve the value of local resource utilization and material supply security.

2 Materials and methods

2.1 Materials

Kukersite-type oil shales (OS) with different kerogen concentrations K-45, K-70 and K-90 (45%, 70% and 90% respectively) were obtained from Oil Shale Competence Center Estonia in Kohtla-Järve. To obtain the kerogen concentrates the OS was milled to under 125 μm and mixed with saline solution to obtain kerogen concentrate of up to 80%, which was washed and dried. To obtain K-90, the concentrate was treated with 2 wt% HNO₃, washed, sieved and dried. The dried OS concentrates were sieved to receive four fractions (63–125 μm; 45–63 μm; <45 μm and <22 μm) and particle size distributions, moisture, metal and elemental composition determined. The fractions were mixed to obtain the required kerogen concentrations. The elemental composition of the oil shales used is shown in Supplementary Table S2 and composition of metals in Supplementary Table S3. The remaining reagents were purchased at the highest purity available and used as received.

2.2 Experimental methods

The screening experiments were carried out in atmospheric and pressurized conditions to study the effects of various parameters (kerogen concentration, time, temperature, oxidant concentration and amount, mixing, pressure, gas composition) on dicarboxylic acid (DCA) yield. The aim was to maximize overall DCA yield whilst retaining the long-chain DCAs (C7-C11). Figure 3 presents schematic illustrations of the atmospheric and pressurized reactor setups used in this work.

The experiments in atmospheric pressure were carried out in 100 mL round-bottom flasks. K-45 or K-70 (2 g) of was weighed into the flask and a specified amount of nitric acid was added. The flask was fitted with a condenser and magnetic stirring at 500 rpm to ensure homogeneous mixing throughout the reaction. Temperature control was achieved and monitored with an internal temperature sensor. The reaction time was recorded from the point at which the internal temperature reached the target value. At the end of the reaction the reaction mixture was cooled to room temperature and collected. The reaction mixtures were centrifuged to remove the remaining solid. The aqueous phase was acidified and, if necessary, an inorganic salt was added to increase ionic strength, followed by multiple extractions with an organic solvent to recover the target organic acids. The solvent was removed, and the samples were esterified using a Dean−Stark apparatus. with a 2:1 methanol−chloroform mixture and a catalytic amount of concentrated H₂SO₄. The dimethyl esters (DMEs) of the DCAs were analyzed with gas chromatography flame ionization detector (GC-MS) to determine the yield.

The elevated pressure experiments were conducted in a modified Parr reactor (4566C, Parr Instrument Company, Moline, IL, United States) with an acid resistant Hastelloy lining. For one set of reactions PTFE lining was tested. The reactor had a bottom-mounted magnetic stirrer to ensure effective mass transfer. A pressure gauge was used to measure the pressure in the reactor, and a manual throttle was used to control the pressure in the reactor. K-45, K-70 and K-90 (2–3 g) were weighed into the reactor, and a specified amount of nitric acid was added. The reactor was monitored with an internal temperature sensor and was heated using a thermostat-controlled heating coil. Beginning of the reaction time was recorded from when the reaction temperature had reached the target value. The gases generated during oxidation resulted in pressure build-up in the reactor. The pressure was either allowed to reach its maximum value, which was recorded, or the throttle was used to partially release the evolved gas to reach target pressures. At the end of the reaction the mixing was stopped and reactor cooled to room temperature. The pressure was released and for some experiments the gas was collected and analyzed with gas chromatography thermal conductivity detector (GC-TCD). The authors were aware of the potential hazards of the evolved gases and applied appropriate precautions during depressurization and gas handling. The oxidation mixtures were collected, and dicarboxylic acid concentrations were determined by capillary electrophoresis without prior esterification. The results were in good agreement with those obtained from esterified samples analyzed with GC-FID.

Along with the nitric acid dosage expressed as a mass ratio, the HNO₃/TOC ratio was used to enable better comparison of the results obtained for different kerogen concentrations. The HNO₃/TOC ratio was calculated as the molar ratio of nitric acid to organic carbon in the kerogen concentrate.

H N O 3 T O C m o l m o l = m H N O 3 g / M H N O 3 g m o l m k e r o g e n c o n c e n t r a t e g × T O C k e r o g e n c o n c e n t r a t e g / g / M C g m o l

2.3 Analytical methods

The dimethyl ester samples were analyzed using a Shimadzu GC-2010 FID gas chromatograph with thermal conductivity detector, equipped with a 30 m polyimide-coated fused silica Zebron ZB-5HT capillary column (0.25 μm film thickness). The capillary column temperature was raised from 50 °C to 300 °C at a rate of 10 °C/min.

The dicarboxylic acid samples were analyzed with capillary electrophoresis Agilent 7100 CE System (Agilent Technologies, Waldbronn, Germany) equipped with diode array detector. Analytes were detected with indirect UV at 380 nm. Detection was performed by indirect UV absorbance at 380 nm. Separations were carried out in fused-silica capillaries (50 μm I.D., 350 μm O.D., total length 70 cm, effective length 61.5 cm). The background electrolyte (BGE) consisted of 15 mM maleic acid and 0.002% HDMB in Milli-Q water, adjusted to pH 11. Samples were injected hydrodynamically at 50 mbar for 8 s. Electrophoretic separations were conducted at 25 °C under an applied voltage of −25 kV. This method also allowed the quantification of acetic and formic acid.

Gas analysis was performed using a Shimadzu GC-2014 gas chromatograph. O₂, N₂, and CO were analyzed using a 3 m molecular sieve 5 Å packed column (OD 1/8″, ID 2 mm, 80/100 mesh), while CO₂ and N₂O were analyzed using a 9 m 25% DC-200/Shimalite packed column (60/80 mesh, OD 1/8″, ID 2 mm). The oven temperature was programmed from 60 °C (holding isothermally for 2 min) to 120 °C (holding isothermally) for 4 min at a rate of 10 °C/min. Although the NO_x peak was observed, its composition (NO/NO₂ ratio) and therefore direct quantification was not possible. Therefore, the NO_x concentration was estimated by difference from the total gas composition.

2.4 Determination of products

In the experiments where derivatization was required for quantification, dimethyl ester (DME) yield was calculated as a percentage with respect to the kerogen content of the raw material. It should be taken into account that esterification adds two carbon atoms to each dicarboxylic acid molecule, resulting in an apparent increase in the reported yield, while not affecting the relative distribution of the acids. In experiments, where the dicarboxylic acids (DCA) were measured directly, the yield of DCAs was calculated as a percentage with respect to the kerogen content of the raw material.

D M E o r D C A Y i e l d % = m a s s o f D M E o r D C A m g m a s s o f k e r o g e n m g × 100 %

The yield of long-chain dicarboxylic acids or their esters is expressed as a percentage relative to the kerogen content of the raw material.

C 7 − C 11 D M E o r D C A Y i e l d % = m a s s o f D M E o r D C A C 7 − C 11 m g m a s s o f k e r o g e n m g × 100 %

The yield of esterified products is shown as a percentage of the total organic solvent extract, after esterification and concentration, relative to the total mass of kerogen in the raw material.

E s t e r i f i e d S a m p l e Y i e l d % = m a s s o f e s t e r i f i e d p r o d u c t m g m a s s o f k e r o g e n m g × 100 %

Oxidation residue was calculated as a percentage with respect to solid content of the raw material, after washing and drying.

O x i d a t i o n r e s i d u e % = m a s s o f s o l i d r e s i d u e m g m a s s o f k e r o g e n c o n c e n t r a t e m g × 100 %

To evaluate conversion, residual organic carbon (TOC) was calculated as a percentage of TOC in the solid residue with respect to TOC in the kerogen concentrate.

R e s i d u a l T O C % = m a s s o f s o l i d r e s i d u e m g × T O C r e s i d u e % m a s s o f k e r o g e n c o n c e n t r a t e m g × T O C k e r o g e n c o n c e n t r a t e % × 100 %

The amount of gas was estimated from the reactor volume, the volume of the liquid–solid products, and the system pressure. Using the measured gas-phase volume composition, the total quantity of each gas could be calculated.

3 Results and discussion

3.1 Atmospheric pressure experiments

3.1.1 Effect of temperature and kerogen concentration

Initial screening experiments with HNO₃ were carried out to evaluate the effect of temperature and kerogen concentration on total DCA, long-chain DCAs (C7-C11), and total esterified sample (mono-, di-, tri-carboxylic and other esters) yield. Nitric acid (1:10 OS to 65 wt% HNO₃ mass equivalent) was added to keep a constant concentration of oxidant in solution. The results are presented in Figure 4.

The results in Figure 4 show that increasing both temperature and reaction time increased the yield of DCAs and solubilized esterified fraction for K-45 and K-70. Approximately 65% of the esterified fraction was DMEs. Temperature had the strongest influence on DCA yield, particularly at shorter reaction times (3 h and 6 h). At longer reaction times (24 h and 48 h), the influence of temperature on DCA formation became less pronounced for K-70. It is known that higher temperatures accelerate the decomposition of nitric acid, promoting the formation of oxidation-initiators, such as HNO₂ and NO₂ (Castellan et al., 1991). This enhances the oxidative degradation of the kerogen matrix and its conversion to dicarboxylic acids (DCAs).

Despite the higher kerogen content in K-70, no significant difference in DCA yields was observed at short reaction times, likely due to excess oxidant availability. Particularly in the case of K-45, the carbonate minerals, quantified with TIC measurement, react with nitric acid through nitrates and CO₂ formation. The sulfide in pyrite in the sandy-clay fraction also consumes nitric acid and forms nitrogen oxides (NO and NO₂), that may initiate secondary oxidation reactions (Meshram et al., 2015; Niidu et al., 2022; Teimouri et al., 2022). Based on TIC and Fe wt% values in Supplementary Tables S2, S3, it can be estimated that carbonates and pyrite consume approximately 8% and 4% of the initial nitric acid for K-45 and K-70, respectively, assuming complete reaction of carbonates to nitrates and sulfide to sulfate. The resulting HNO₃/TOC molar ratios can be calculated as 3.3 for K-45 and 2.2 for K-70. At longer reaction times (24 h and 48 h) and milder temperatures (80 °C and 100 °C), K-70 exhibited greater yields than K-45, showing that the lower HNO₃/TOC molar ratio did not limit the formation of DCAs. The higher yields for K-70 may reflect reduced nitric acid consumption by the mineral fraction. However, at 120 °C, where nitric acid thermal decomposition is most pronounced, K-45 produced higher DCA yields than K-70 throughout, possibly due to the greater molar excess. These results indicate that extensive beneficiation prior to processing is unnecessary and lower-grade OS can serve as a viable feedstock for DCA production.

The results in Figure 4 also show that the yield of longer DCAs formed during oxidation increases with overall DCA yield as expected and plateaus over time. However, extended reaction times and increased temperatures, most pronounced at 120 °C, show a small decrease in longer DCA yield, which could be attributed to the decomposition of long-chain DCAs in harsher conditions over time.

3.1.2 Effect of oxidant concentration and amount

To improve safety at larger scale and limit oxidant amount in the system, the effect of HNO₃ concentration and oxidant amount on yield was screened with K-45. Temperature of 120 °C was chosen as increased temperatures showed greater overall yields, especially at shorter reaction times. Nitric acid with 65 wt% or 30 wt% concentration was added in a 1:10 or 1:5 OS to HNO₃ mass equivalent and oxidation was carried out for 3–24 h. The HNO₃/TOC molar ratio varied from 3.4 to 1.6 at 65% concentration and 1.5 and 0.64 for 30% concentration. The results are presented in Figure 5.

Higher concentration and mass equivalent of HNO₃ relative to OS resulted in higher DCA yields. Nitric acid concentration had a greater influence on yield than the HNO₃/TOC molar ratio. Higher yields were obtained using a 1:5 ratio of more concentrated acid compared to a 1:10 ratio at 30% concentration, despite similar HNO₃/TOC molar ratios. Higher nitric acid concentrations are less thermally stable and more prone to decomposition to oxidation-initiating species such as HNO₂ and NO₂, which can lead to radical formation and oxidation, resulting in greater yields at shorter reaction times. At longer reaction times (24 h), 1:10 ratio with 30% nitric acid achieved yields comparable to those obtained at shorter times using 65% nitric acid, showing that extended reaction time can somewhat compensate for lower oxidant concentration. This indicates that DCA formation is primarily kinetically controlled rather than limited by the total oxidant stoichiometry at these conditions.

The yield of longer DCAs correlated with increased total yield at 3 h and 6 h for both concentrations. After 24 h, the yield of the longer DCAs had plateaued and decreased slightly for 65% concentration, whilst for 30% concentration their yield had increased significantly. This indicates that milder conditions, such as lower concentrations of oxidant can give comparable yields of longer DCAs over extended reaction times. Higher nitric acid concentrations are needed to obtain higher yields of shorter DCAs but might lead to the decomposition of longer-chain acids under more aggressive oxidative conditions over time.

The effect of oxidant amount was also evaluated for K-70 at a lower temperature (80 °C) and extended reaction time of 48 h. 65 wt% HNO₃ was added in varying amounts, corresponding to HNO₃/TOC molar ratios of 2.3, 1.1, 0.69, and 0.46. The results are shown in Table 1.

TABLE 1. Total and long-chain DME yield with varying oxidant amount.

After 48 h, similar yields were obtained for the 1:10 and 1:5 OS to HNO₃ mass ratios, whereas lower yields were observed for the 1:3 and 1:2 ratios. This suggests that increasing oxidant availability enhances DCA formation up to a certain point, beyond which further increase has little effect. The same trend was observed for longer DCAs, with yields reaching a plateau at an OS/HNO₃ ratio of 1:5. Despite the longer reaction time, overall DCA yields at 80 °C remained lower than those achieved at 120 °C, highlighting the importance of elevated temperatures for faster overall reaction kinetics. However, for the 1:10 and 1:5 mass ratios, the yield of longer DCAs reached a similar maximum of approximately 9.5% at both temperatures, suggesting that their formation is less temperature-dependent and proceeds with relatively rapid kinetics under sufficient oxidant availability.

3.1.3 Effect of sulfuric acid

The influence of H₂SO₄ addition on oxidation was screened with K-45. Sulfuric acid was added in amounts equivalent to the inorganic fraction (either 100% or 200%). The addition of H₂SO₄ was investigated for two principal purposes: to bind the inorganic carbon present in the oil shale matrix and to assess its potential catalytic effect on the oxidation reaction. The oxidation tests were conducted at a temperature of 120 °C with a 1:5 and 1:10 OS to 30 wt% HNO₃ mass ratio. The concentration of HNO₃ of 30% was selected to limit nitration potential. The results are shown in Figure 6.

The addition of H₂SO₄ had a significant effect on yield at the lower 1:5 mass ratio. Sulfuric acid reacts with the inorganic fraction, especially carbonates, which leads to the greater effective availability of nitric acid to oxidize the organic material. The higher 1:10 ratio, did not show a large difference in the yield in comparison, as the oxidant is present in greater molar excess.

Further experiments with sulfuric acid were not conducted, as the addition increases the risk of nitration. Moreover, from a scale-up perspective, the use of a mixture of two strong mineral acids would complicate process design, as well as the handling and treatment of process streams and by-products.

From atmospheric pressure experiments, it could be concluded that temperature and oxidant concentration had the largest effect on yield. To achieve high yields at short reaction times, both temperature and oxidant concentration should be increased to promote the formation of oxidation-initiators, such as NO₂ and NO, leading HNO₂ and radical formation and oxidation. This approach increases the yield of shorter DCAs, but over extended reaction times promotes the decomposition of long-chain DCAs. To maximize the proportion of the long-chain acids and reach good overall yield, milder conditions such as lower temperatures and reduced oxidant concentrations and amounts should be applied over extended reaction times. The effect of kerogen concentration did not have a significant effect on yield.

3.2 Elevated pressure experiments

Further screening experiments at elevated pressures up to 38 bar were conducted in Parr reactors with the aim of reducing reaction time and oxidant amount. It is known that as pressure is increased, the solubility of NO and NO₂ is improved, especially in aqueous conditions (Ting et al., 2013). Higher dissolved NO_x concentrations enhance the rate of formation of nitrous acid (HNO₂) and other oxidation intermediates, leading to radical formation and oxidation. Radical-mediated oxidation reactions are believed to dominate under high-temperature reaction conditions for cyclohexanol oxidation (Castellan et al., 1991). Although, the direct measurement of NO and NO₂ concentrations and NO_x absorbed was not possible, a thorough evaluation of the key parameters influencing DCA yield was carried out.

3.2.1 Effect of temperature, time and mixing

A three-factor, two-level design of experiments with one center point was implemented to assess the influence of time, temperature, and mixing on oxidation. The experiments were performed in a randomized order. For each experiment K-90 (2 g) was reacted with 30 wt% HNO₃ (45 g) at a constant HNO₃/TOC molar ratio of 1.8. The pressure in the reactors was kept at 5 bar and excess pressure was released through a valve. The results are presented in Table 2.

TABLE 2. Total and long-chain DCA yield with varying temperature, time and mixing speed for K-90 at 5 bar.

The increased pressure had a positive effect on DCA yield, as yields at around 21% were reached in 2 h, compared to 9.8% DME yield at atmospheric pressure after 3 h with similar HNO₃/TOC molar ratio of 1.6% and 30% concentration as shown in Figure 5.

The data were analyzed using the fit model function on JMP. Different models were fitted and analyzed by determination coefficient (R²), adjusted determination coefficient (R² _adj) and p-values of the model, as well as variance (ANOVA) and lack-of-fit. Mixing showed no significant effect on yield and thus was removed as a parameter from the model. The best model gave an R² value of 0.983 and R² _adj of 0.974 with multiple linear regression with the equation:

D C A y i e l d % = − 58.38 + 0.585 T + 0.0694 τ + 0.0037 T τ

Overall, temperature had the greatest effect on yield (0.585 coefficient), followed by time (0.0694 coefficient) and Tτ (0.0037 coefficient). The model gave a mean response of 12.24 with root mean square error of 1.02. This corresponds to an average prediction error of 8%, which fits the error in the repeated measurements. The ANOVA results are presented in Supplementary Table S4. The low p-value (<0.0001) showed that the variables had a significant effect on the response (yield). The p-values for each parameter were under <0.05, specifically <0.0001 for temperature (T) and time (τ) and 0.00269 for the two-way interaction (Tτ) showing the significance of each parameter on the response, yield. Lack of fit test, displayed in Supplementary Table S5 was not significant (p = 0.7045), indicating that the model adequately fits the experimental data.

The results are consistent with the results obtained under atmospheric conditions, where higher temperatures led to increased yields. In pressurized reactors, temperature exerts a substantially greater effect than time, indicating faster reaction kinetics under elevated pressure. While the reaction mechanism was not investigated directly, this could be attributed to increased solubility of the reaction gases, including nitrogen oxides (NO and NO₂) and the improved gas–liquid-solid mass transfer under elevated pressure. This could increase the effective concentration and availability of oxidation-active species and increase the reaction rates.

3.2.2 Effect of temperature, oxidant concentration and amount

A three-factor, two-level design of experiments with one center point was implemented to investigate the effect of oxidant concentration and amount on yield at different temperatures with the reaction time of 1 h. One repetition experiment was conducted. The HNO₃/TOC ratios were calculated and ranged from 0.33 to 2.0. Maximum pressure for each reaction was recorded. The results are presented in Table 3.

TABLE 3. Total and long-chain DCA yield and maximum pressure with varying temperature, oxidant concentration and amount for K-70.

Similarly to the results from atmospheric experiments, increased HNO₃ concentrations had the greatest effect on yield. Higher concentration and oxidant amount, as well as temperature resulted in greater pressure build-up in the reactor (P_max). The increased pressure improved yields, as 19.9% yield was reached in 1 h at 23.6 bar, compared to about 15% yield at 5 bar and 9.8% DME yield at atmospheric pressure at similar conditions, as seen in Table 2 and Figure 5. These results suggest that increased pressure promotes DCA formation, enabling higher yields at lower temperatures and oxidant amounts.

The data were analyzed using the fit model function on JMP using the same methodology as described above. Pressure and HNO₃/TOC molar ratio were also tested as model parameters but did not improve the fit and were therefore excluded from the final regression. The best model gave an R² value of 0.946 and R² _adj of 0.919 with multiple linear regression with the equation:

D C A Y i e l d % = − 56.88 + 0.732 c + 0.529 m + 0.388 T

Oxidant concentration had the greatest influence on yield (0.732), followed by oxidant amount (0.529) and temperature (0.338). The model gave a mean response of 15.59 with root mean square error of 2.72 corresponding to a prediction error of 17%, which is less accurate than the previous model. The results of ANOVA are shown in Supplementary Table S6. The low p-value (0.0003) showed that the variables had a significant effect on the response (yield). The p-values for each parameter were under <0.05, specifically 0.0002 for oxidant concentration (c), 0.0282 for oxidant amount (m) and 0.0055 for temperature (T) showing the significance of each parameter on the response, yield. Furthermore, the lack of fit test presented in Supplementary Table S7, was not significant (p = 0.1475) indicating that the model adequately describes the experimental data.

The results are consistent with those obtained under atmospheric conditions, where increasing the oxidant concentration had the most pronounced effect on yield and a sufficient excess of oxidant (HNO3/TOC ≈ 0.67) was required to achieve higher yields. Increasing the pressure above 5 bar had a positive effect on yield. Although the effect of temperature was less pronounced than that of oxidant concentration and amount, it remained significant. Increased temperatures can partially compensate for lower oxidant concentrations, which is advantageous from an industrial perspective for reducing oxidant consumption.

3.2.3 Effect of pre-pressurization and HNO₃/TOC ratio

The effects of HNO₃/TOC molar ratio and reactor pre-pressurization to 10 bar on yield and product gas composition were investigated for 1 h reaction times at 130 °C. In each experiment K-45 (3 g) was reacted with 30 wt% HNO₃, corresponding to HNO₃/TOC molar ratios of 1.5, 2, 3 and 4. To evaluate the influence of initial reactor pressure and additional oxygen, the reactors were flushed and pre-pressurized to 10 bar with either N₂ or an N₂/O₂ (50:50) mixture, after the addition of K-45 and HNO₃, prior to heating. The maximum pressure was recorded before depressurizing. The results are shown in Table 4. The use of excess oxidant above HNO₃/TOC molar ratio of 3:1 did not increase the yield. Pre-pressurizing the reactor with N₂ or N₂/O₂ had no significant effect on DCA yield, CO₂ or N₂O formation. Although NO_x concentrations were not measured directly, no NO_x was detected in the reactor off-gas following pre-pressurization with an N₂/O₂ atmosphere at HNO₃/TOC ratios above 1.5. This could be explained by the oxidation of NO to NO₂ by oxygen and the subsequent hydrolysis back to nitric acid. However, as nitrogen oxides are required to form HNO₂ and reactive radicals in solution that drive the oxidation, regeneration of nitric acid alone did not have a positive effect on yield.

TABLE 4. Total DCA yield and gas composition with varying HNO₃/TOC molar ratio and pre-pressurizing conditions.

3.2.4 Effect of pressure

The effect of pressure on DCA yield and gas composition was studied at 140 °C for 30 min. For each experiment K-45 (3 g) was reacted with 30 wt% HNO₃, corresponding to HNO₃/TOC molar ratio of 3. A throttle was used to maintain the reactor pressure at target value. At the end of the reaction, the reactor was cooled to room temperature and a sample for gas analysis was collected. The results can be found in Table 5. The gas sample bag had leaked for the reaction at 5 bar.

TABLE 5. Total DCA yield and gas composition with varying reactor pressure.

Increasing pressure from 5 to 15 bar resulted in an increase in DCA yield, reaching a maximum of 30% at 15 bar. However, at pressures above 15 bar the increase of pressure did not further increase the yield. The gas composition showed a small increase of CO, CO₂ and N₂O and decrease in the estimated NO_x concentrations with increasing pressure, indicating deeper oxidation depth. This suggests that elevated pressures might lead to the increased solubility and availability of oxidation-initiating nitrogen oxides (NO and NO₂), which in turn promote HNO₂ and radical formation and oxidation.

3.2.5 Material selection

The reactor was fitted with a PTFE lining as an alternative to Hastelloy to minimize the risk of corrosion under the applied reaction conditions. Oxidations of K-45 at further elevated temperatures of 140 °C were conducted, using 30 wt% HNO₃ in a HNO₃/TOC molar ratio of 3:1 in both reactors. The material of the reactor did not have a significant effect on yield or conversion, expressed as oxidation residue as can be seen in Figure 7.

To see whether titanium would be a suitable material for the reactor, a piece of titanium was placed in the reactor and oxidations of K-45 at temperatures of 130 °C and 140 °C were conducted for 1 h, using 30 wt% HNO₃ in a HNO₃/TOC molar ratio of 3:1. The addition of titanium seemed to have a positive effect on DCA yield, as seen in Table 6.

TABLE 6. Total DCA yield in Hastelloy- and PTFE-lined reactors with titanium piece.

The reason for this improvement is currently unclear but may be related to catalytic or other surface-mediated effects. However, the titanium surface was not characterized before or after the experiments nor was its corrosion behavior investigated. Previous studies have shown that titanium was the most durable material to nitric acid oxidation conditions and was a suitable material for reactors and welding (Veski and Veski, 2019). The observed increase in DCA yield requires further investigation, but titanium may be considered as a potential reactor material or lining for future reactor designs.

3.2.6 Reactions in Hastelloy- and PTFE-lined reactors over time

Distribution of organic carbon between the solid, liquid and gas phase was studied over time for K-45 at 140 °C and 30 wt% concentration of nitric with HNO₃/TOC molar ratio of 3:1. The reactions were performed in Hastelloy- and PTFE-lined reaction vessels. The results for the PTFE-lined reactor are shown in Figure 8.

The results indicate that carbon conversion from the solid phase approached completion after approximately 40 min. Similarly, DCA yield increased over time, reaching a maximum at 30–40 min. The yield of longer-chain DCAs was highest at 30 min and decreased thereafter, probably due to decomposition. As residual kerogen (TOC) became depleted, further reaction mainly resulted in increased CO₂ and CO formation. A similar trend is followed in Hastelloy reactors. Consequently, reaction times of approximately 30–40 min appear optimal for achieving higher yields while preserving a larger fraction of longer-chain acids.

The carbon mass balance for the 40-min reaction in the PTFE-lined reactor can be found in Supplementary Table S8. Overall, 56.2% of the carbon was recovered in the liquid phase, including 14.7% as DCAs, 1.4% as acetic acid, and 40.1% as other dissolved organic compounds. The gaseous products accounted for 43.0% of the carbon, comprising 35.6% as CO₂ and 7.4% as CO, while only 0.9% remained in the solid residue. Compared to WAO followed by subsequent 8% nitric acid oxidation, approximately 24% of the carbon was recovered in the liquid phase, 50%–60% in gaseous products, and 16%–29% in the solid residue (Kaldas et al., 2025). Nitric acid oxidation alone resulted in considerably greater carbon solubilization, lower solid residue formation, and reduced carbon losses to the gas phase.

Acetic acid concentrations were measured over time for PTFE and Hastelloy reactors and the results are shown in Supplementary Figure S1. The oxidative decomposition of phenols and similarly resorcinols results in the formation of low molecular acids such as acetic acid in oxidative conditions. The yield of acetic acid is about 4% and follows similar trends to overall DCA yield.

Nitrous oxide formation was monitored over time for the two reactors, the results are shown in Supplementary Figure S2. Around 2.5% of N₂O was formed per oxidant already at 10 min and remained constant over time. This suggests that the majority of nitric acid is reduced to the stable N₂O fast and reaction time has a small effect on its formation.

3.2.7 Scale-up considerations

Scale-up of kerogen oxidation with nitric acid requires several engineering considerations from the safety, reactor and process design perspective. The reaction is strongly exothermic and based on the results of this study should be carried out at elevated temperatures (140 °C) and pressures (10–15 bar). Consequently, efficient heat removal and temperature control are essential to ensure safe operation at larger scale. A continuous-flow reactor is expected to offer significant advantages over batch processing, including improved heat and mass transfer and reduced temperature gradients and hot spots, thereby enhancing process safety (Capaldo et al., 2023). Based on the findings of the present study, a continuous-flow reactor without mechanical agitation, termed the Kerox process, was designed and constructed. However, the design, operation, and performance of this reactor are beyond the scope of the present work and will be presented in a separate publication.

Under the reaction conditions employed, the elevated temperatures were expected to favor oxidative degradation over nitration or nitrosation. Consistent with this expectation, no volatile nitrated or nitrosated compounds were detected in the extracted and esterified solubilized organic fraction by GC–MS. An example GC–MS chromatogram and the identified compounds are provided in Supplementary Figure S1 and Supplementary Table S9. The oxidized solid residue contained only 0.37 wt% nitrogen after 40 min, by which point the solubilization of kerogen was essentially complete. After rectification of the esterified DCA mixture, the cubic residue, which accounts for ∼10% of the solubilized material, had a nitrogen concentration of 2.5%. The elemental compositions of the oxidized solid residue and cubic residue are shown in Supplementary Table S10. Although these analyses cannot exclude the presence of trace amounts of nitrated or nitrosated species, they indicate that these are not the significant reaction pathways under the conditions investigated. From a process safety perspective, this reduces concerns regarding the accumulation of potentially hazardous nitrogen-containing organic compounds.

Industrial implementation of the process will require appropriate treatment of gaseous and liquid process streams. NO_x containing off-gases should be treated in an absorption scrubber to recover nitric acid for recycle and to reduce overall oxdiant consumption. More importantly N₂O formation should be minimized, as it cannot be recycled back to nitric acid and has a global warming potential 273 times that of CO₂. Several mitigation methods, such as catalytic decomposition of N₂O, may be applied to reduce greenhouse gas emissions (Wu et al., 2024).

To reduce oxidant consumption and improve process safety, the residual nitric acid concentration in the liquid product stream should be minimized. Although vacuum distillation has previously been proposed for nitric acid recovery, the formation of an HNO₃–H₂O azeotrope results in the oxidation products remaining in the concentrated acid phase (Veski and Veski, 2019). In our preliminary studies, this approach resulted in partial decomposition of the oxidation products. Alternative downstream processing strategies, such as liquid-liquid extraction or esterification of dicarboxylic acids followed by distillation of the corresponding diesters, may therefore provide more suitable routes for product recovery and are currently under investigation.

4 Conclusion

Nitric acid oxidation of kukersite kerogen was investigated in atmospheric and pressurized systems by varying nitric acid concentration (20–65 wt%) and kerogen content (45–90 wt%). The aim was to identify conditions enabling direct chemical valorization of kerogen into valuable intermediates, specifically dicarboxylic acids.

Maximum dicarboxylic acid yields of 30% were achieved at 130 °C–140 °C and 10–15 bar within 30–40 min with 30 wt% nitric acid concentration, whereas atmospheric conditions required significantly longer reaction times (24 h) and higher oxidant concentrations (65 wt%) to reach comparable yields. Temperature and nitric acid concentration were the dominant parameters, while oxidant excess had a limited benefit beyond HNO₃/TOC molar ratio of 1. Elevated pressure was shown to increase reaction rates. Product distribution depended strongly on reaction severity: higher overall yields were obtained under harsher conditions, but with increased degradation of longer-chain DCAs (C7–C11).

Overall, nitric acid oxidation provides a viable route for kerogen-to-DCA conversion. The use of a single oxidant simplifies process design and supports scalability, offering a pathway toward chemical valorization of oil shale resources.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors acknowledge the use of OpenAI’s ChatGPT (version: ChatGPT, model: GPT-5.5) in the preparation of the schematic figure presented in this manuscript. The AI-generated figure was reviewed, edited, and refined by the authors, who take full responsibility for its final content, accuracy, and interpretation.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fceng.2026.1880713/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Xianhai Zeng, Xiamen University, China

Reviewed by: Mikhail A. Proskurnin, Lomonosov Moscow State University, Russia

Zihong Xia, East China University of Science and Technology, China