Four outdoor monitoring tanks, each holding 2 m³ of slurry were used to estimate the CH₄ emissions from untreated cow manure (UCM), which served as the non-plasma treated control, and nitrogen enriched organic fertiliser (NEO). All tanks were equipped with a lid that remained open, except during the 40-min measurement period when the lid was closed and gas samples were drawn at 10-min intervals to estimate CH₄ emission. Figure 1 shows a schematic of the system. Continuous measurements of the pH and temperature in all tanks were conducted the first year.

The UCM originated from Skollenborg, Norway and was filtered with a screw press before being transferred to two of the tanks. The same filtered UCM was used as input material in the plasma process, which converts UCM to NEO. Figure 2 illustrates the plasma process and Table 1 shows the difference in chemical composition between the UCM and NEO (averaged over the three years). Each year, the tanks were drained before a new batch of UCM and NEO was transferred to the tanks, with approximately the same composition as the year prior.

In the laboratory experiment, the UCM was compared to UCM acidified with sulfuric acid (H₂SO₄) to pH 6.3 and pH 5.0 (UCM_pH6.3 and UCM_pH5.0) and to NEO at pH 6.3 and pH 5.0 (NEO_pH6.3 and NEO_pH5.0). Each UCM treatment was set up in nine gas tight serum-vials, the NEO treatment in three. Conditions in the vials were made anoxic by washing the bottles (repeated cycles of evacuation and He-filling while stirring the slurries) to accelerate methanogenesis. The vials were incubated at 30°C for 180 h in a water bath connected with a multi-column GC (Model 7890A, Agilent, Santa Clara, CA, USA) and a chemoluminescence NO analyser (Model 200 A, Advanced Pollution Instrumentation, San Diego, USA) via an autosampler and a peristaltic pump for repeated headspace analyses as described by Molstad et al. (2007). The methane concentrations in the headspace were measured six times a day.

Approximately halfway through the incubation, three vials from the UCM and UCM_pH6.3 received a potassium nitrate (KNO₃) dose resulting in a concentration of 1,000 mgNO₃⁻- N/L, and another three vials received a potassium nitrate (KNO₃) and potassium nitrite (KNO₂) dose resulting in a concentration of 1,000 mgNO₃⁻- N/L and 750 mgNO₂⁻- N/L, which corresponds to the nitrogen oxyanion concentration in NEO. No CH₄ formation was detected in the UCM_pH5.0 vials. Therefore, no NO₃⁻ or NO₂⁻ was added to the UCM_pH5.0 treatments. The setup aimed to determine whether pH, nitrate or nitrite, or a certain combination of the three, yielded a complete inhibition of methanogenesis or whether the inhibition is a result of particular properties of the plasma treatment.

Temperature readings suggested differences in microbial activity between the two treatments. As shown in Figure 3, the temperature at a slurry depth of 30 cm displayed higher peaks in UCM than in NEO, and more pronounced diurnal fluctuations, likely caused by higher microbial activity in the UCM that affected the slurry temperature, whereas air temperature was the main driver in NEO. However, the minima are lower in the UCM – resulting in comparable average temperature over the entire storage period.

The pH during storage also indicated a clear inhibition of microbial activity in NEO (Figure 4). The pH in NEO rose slowly from pH 5.3 to pH 5.9 throughout the 80 days of storage. The slow increase can be attributed to deprotonation of carbonates, combined with the reaction of NO₂⁻ with organic matter. The onset of microbial denitrification would have been associated with an abrupt pH increase and elevated temperatures as described by Hiis et al. (2023), suggesting an absence of denitrification in the stored NEO.

The apparent inhibition of microbial activity had a clear effect on the CH₄ fluxes. Figure 5 shows the cumulative CH₄ emissions for all tanks in the 2021–2023 measurement season, where the methane formation is completely inhibited by the nitrogen enrichment. A minimal yet consistent CH₄ uptake is found in all NEO treatments.

The low measurement frequency in 2022 did not yield an appropriate resolution of the fluxes, especially since the deviation in emissions between the two tanks in 2021 was large, mainly as a result of crust formation on UCM-1 and no crust formation on UCM-2. CH₄ emissions were substantially higher in 2023 compared to 2022, a finding we currently have no explanation for, other than speculating that the annual temperature variations may have influenced the rate of CH₄ formation. Despite emission-variation in the UCM, the reduction resulting from the plasma-based nitrogen enrichment was reliably above 100% throughout the storage periods. The complete emission reduction distinguishes the plasma-based nitrogen enrichment from acidification alone, which reports CH₄ reductions in the range of 69 to 84% (Habtewold et al., 2018).

The controlled incubation experiments demonstrated a pronounced effect of acidification and NO₃⁻ and NO₂⁻ addition on CH₄ production, although their combinatory effect did not match that of the plasma treatment entirely. The unacidified UCM (pH 7.1) showed a steady methane formation, until either NO₃⁻ or NO₃⁻ + NO₂⁻ was added (Figure 6). The addition of NO₃⁻ + NO₂⁻ resulted in inhibition of methanogenesis throughout the incubation period, whereas the methane formation resumed in the NO₃⁻ treatment.

Both treatments where the nitrogen oxyanion concentration was artificially elevated resulted in denitrification, as was evident from N₂ formation shown in Figure 7. The pronounced drop in CH₄ concentration was unexpected; the increased formation of N₂, N₂O, NO and CO₂ and could not entirely explain the reduced concentration. A potential explanation of the effect is denitrifying anaerobic methane oxidation (DAMO), as described by Wei et al. (2022), a thermodynamically favourable denitrification pathway where CH₄ acts as electron donor and is oxidised to CO₂.

Figure 8 shows the impact of acidifying the UCM with H₂SO₄ to pH 6.3. Acidification alone resulted in a 45% reduction in CH₄ accumulation. The nitrogen oxyanions were added in identical concentrations as in the prior experiment after 86 h. Both treatments resulted in complete inhibition of methanogenesis. NEO (UCM plasma treated to pH 6.3) showed a complete inhibition of methanogenesis throughout the experiment.

Both UCM_{pH6.3 + NO3} and UCM_{pH6.3 + NO3 + NO2} exhibited clear signs of denitrification. The pH at the end of the experiment was 8.1 and 8.2, which indicates that the alkalization occurred due to consumption of nitrogen oxyanions. In contrast, pH in NEO rose moderately from pH 6.3 to pH 6.5 and did not display the same N₂O peaks as the comparable UCM treatments, as shown in Figure 9.

When acidifying the UCM to pH 5.0, no methane formation occurred over the duration of the experiment. However, the amount of H₂SO₄ required to reach this pH results in a sulphur content that often exceeds agronomic needs (5.4 kg 98%H₂SO₄ m⁻³). Targeting such low pH levels with sulfuric acidification should therefore be avoided.

The exact inhibitory mechanisms of plasma-based nitrogen enrichment cannot be determined by this set of experiments alone. However, some important observations can be made: the addition of nitrogen oxyanions suppresses CH₄ formation, likely because of a shift to NO₃⁻ and NO₂⁻ as the favoured electron acceptors. Consequently, denitrifiers consume the nitrogen oxyanions whereafter methanogenesis is likely to resume (as in UCM_NO3). Our results also indicate the occurrence of the DAMO process, as denitrification coincides with a reduction in the CH₄ concentration that cannot be accounted for by dilution alone. Surprisingly, the NEO, even at a moderate pH of 6.3, did not result in the denitrification that occurred in its chemical imitations, suggesting that the plasma treatment itself has antimicrobial properties beyond the acidification and addition of nitrogen oxyanions. One hypothesis worthwhile exploring is that the plasma treatment occurs under aerobic conditions when the facultative anaerobic denitrifiers are more likely to respire oxygen and therefore cannot easily counter the elevated levels of undissociated nitrite (HNO₂).