Biodesulfurization of the mixture of dibenzothiophene and its alkylated derivatives by Sphingomonas subarctica T7b

Organosulfur compounds classified as dibenzothiophenes (DBTs) and their derivatives are contained in petroleum. When used as fuel, these substances release SOx emissions, thus contributing to air pollution, acid rain, and climate change. Therefore, it is necessary to reduce the content of these organic sulfur compounds in fuels and one way to achieve this is through bacterial desulfurization. This study reports the biodesulfurization process of a mixture of DBT, 4‐hexyl DBT, 4,6‐dibutyl DBT, and various organosulfur compounds in light gas oil (LGO). The experiment was conducted by treating 1 mL of aromatic organosulfur compounds with 100 mg/L in n‐tetradecane or 1 mL LGO with 5 mL mineral salts in sulfur‐free medium, incubated at 27 °C for 5 days with shaking at 273 rpm. Gas chromatography analyses revealed that the growing Sphingomonas subarctica T7b cells desulfurized and converted 88.29% of DBT to 2‐hydroxybiphenyl as a metabolite while a mixture of DBT and 4,6‐dibutyl DBT was desulfurized at 86.40% and 7.00%, respectively. Furthermore, the mixture of DBT, 4‐hexyl DBT, and 4,6‐dibutyl DBT had a desulfurization percentage of 84.40%, 41.00%, and 6.66%, respectively, after five days of incubation. The compounds were observed to desulfurize slightly better as single compounds compared to when mixed with other aromatic sulfur compounds.


Introduction
Petroleum is a crucial nonrenewable energy source world wide, with high demand from numerous industries (Bor doloi et al. 2014; Gunam et al. 2020. However, the com bustion process is known to produce hazardous wastes. This includes high sulfur content, which is unfriendly to humans and the environment. These sulfuroxides (SO x ) emissions produce air pollution, acid rain and contribute to climate change (Kobayashi et al. 2001; Gunam et al. 2006; Etemadi et al. 2018. Oldfield et al. (1997) identified at mospheric SO x as a major factor in urban environmental air pollution, alongside acid rain, known to play a signifi cant role in forest decline globally.
Light gas oil (LGO) is classified as crude oil and is treated by hydrodesulfurization (HDS). After this process, a few hundred parts per million of sulfur are usually left, and these include dibenzothiophene (DBT), benzothio phene (BT) and their derivatives. According to Zhang et al. (2013), complex alkylDBT mixtures in crude oil tend to be altered by individual microbial biodesulfuriza tion. Hence, HDS has been routinely carried out in re fineries as an alternative to reduce petroleum sulfur con tent (Bhatia and Sharma 2010). This process uses metal complex chemical catalysts as well as extremely high tem peratures and pressures (Sohrabi et al. 2012; Gunam et al. 2013. However, the procedure is ineffective for the par ticular polyaromatic sulfur heterocycles (PASHs) present in heavier fractions (Mohebali and Ball 2016). This has mandated further research on biodesulfurization as an al ternative technology for a more efficient PASH degrada tion (Sohrabi et al. 2012; Gunam et al. 2021).
DBTdesulfurizing microorganisms with cleavage ac tivity of CS bonds are expected to effectively biocatalyze petroleum biodesulfurization and this reaction uses a '4S' pathway, refers to the four intermediates formed (DBT sul foxide, DBT sulfone, hydroxyphenyl benzene sulfonate, sulfite), and this pathway occurs through successive oxi dation of DBT (Boniek et al. 2015; Yi et al. 2019. Gunam et al. (2006) previously reported that Sphingomonas sub arctica T7b is able to desulfurize alkyl DBTs and BTs in this manner.

Materials
The reagents used in this experiment include, DBT from Tokyo Kasei Kogyo Co. Ltd., alkyl DBTs from Nard Insti tute Ltd. (Hyogo, Japan), organic solvent (ntetradecane) from Wako Pure Chemical Co., Osaka, Japan, and LGO from Petroleum Energy Center, Shizuoka, Japan. All other reagents met analytical standards and therefore did not require further purification before use. Concentrated aromatic compounds fraction (CA) was obtained through commercial LGO fractionation (Gunam et al. 2006).

Bacterial strains and culture media
Sphingomonas subarctica T7b was isolated from soil sam ples contaminated with oil through enrichment culture, as described by Gunam et al. (2006). Meanwhile, the biodesulfurization evaluations were performed in a min eral salt sulfurfree (MSSF) medium and an organic n tetradecane, with 10 g/L glucose as a carbon source, sup plemented by 0.54 mM (100 mg/L) DBT as a sole sul fur source. The enrichment culture (MSSFCA) com prised MSSF, ethanol and CA in the ratio 1000: 2:1 (v/v), while the desulfurization assessment agent (MSSFTD) contained MSSF medium and ntetradecane in the ratio 5:1 (v/v), with various sulfur compounds dissolved in the or ganic solvent.

Seed culture preparation and bio-desulfurization analyses
For this preparation, fresh 5 mL of MSSFCA medium was inoculated with 2% (v/v) of midlog phase of S. subarc tica T7b preculture and incubated at 27°C for 4 d with 273 strokeper minute reciprocal shaking. Meanwhile, 6 mL of the desulfurization assay medium (5 mL of MSSF and 1 mL of TD) was also inoculated with 0.1 mL of seed culture (OD 660 of 5) and cultured under similar conditions. Subsequently, the sulfur compound infused ntetradecane organic layer and the water layer (MSSF medium) were separated through centrifugation at 20,000 × g for 10 min at 4°C. This process was repeated for the control sam ples (uninoculated medium), as described by Gunam et al. (2006).

Analytical procedures
Bacterial growth and cell concentration were evaluated us ing water layer OD 660 measurements (Gunam et al. 2006), while dibenzothiophene and DBT derivatives were mea sured through flame photometric detector gas chromatog raphy (GCFPD). Similarly, the concentrations of these compounds in growth culture were analyzed by GC14A (Shimadzu, Japan) equipped with a column Zebron ZB1 (60 m × 0.25 mm × 0.25 μm; Phenomenex, USA) and an FPD with an initial column oven temperature of 220°C , heated to 280°C at a rate of 3°C min 1 , was utilized. The equipment's injection and detector temperatures were maintained at 300°C. Subsequently, DBT and derivatives as well as the metabolite 2hydroxybiphenyl (2HBP) con centrations were measured with a gas chromatography flame ionization detector (GCFID). Meanwhile, the sam ples were acidified to pH 2 with 6N HCl and extracted with 0.5 volumes of ethyl acetate. An aliquot of this ex tract was centrifuged and GCFID (GC17A, Shimadzu) equipped with a DB17 column (0.25 mm i.d. × 30 m length; J&W Scientific, Folsom, CA), was used to ana lyze the supernatant. Helium was utilized as the carrier gas for this analysis and the injector as well as detector temperatures were set at 260°C. The 2HBP metabolite was identified with gas chromatographymass spectrom etry (GCMSQP5000; Shimadzu), while the organosul fur compound distribution in light gas oil was analyzed by GCatomic emission detector (GCAED) (Yu et al. 2006). Figure 1 shows the T7b strain has the ability to grow on MSSFTD media containing organosulfur compounds and desulfurize DBT, 4,6dibutyl DBT, and 4hexyl DBT, re spectively at 82.23; 8.27; and 39.97% after cultivation at 27°C with shaking at 273 rpm for 5 d. DBT compounds FIGURE 1 S. subarctica T7b's growth as well as desulfurization activity in MSSF medium and n-tetradecane as a model oil containing the single substrate of DBT and its derivatives (100 mg/L), incubated at 27°C, and shaken at 273 rpm for 5 d.  are therefore much easier to desulfurize and separate com pared to the other two derivatives. In addition, the mi croorganism's cell growth rate was discovered to be di rectly proportional to the ability to desulfurize the com pounds tested. Figure 2 shows the growth and desulfurization obtained from growing S. subarctica T7b cells in MSSFTD medium with DBT as the sole sulfur source at 27°C. Af ter 120 h cultivation, the cells of T7b strain OD 660 reached 2.17, while the residual DBT and 2hydroxybiphenyl (the product of the '4S' biodesulfurization pathway) concentra tions were 14.88 mg/L (88.29% desulfurization) and 69.86 mg/L, respectively. The DBT growth and desulfurization patterns exhibited were similar to the previous study coun terparts (Gunam et al. 2006). According to GCMS anal ysis, this product has the same mass spectrum as authentic 2HBP, and DBT disappeared from the medium faster than 2HBP accumulated. Figure 3 shows that in a mixture of two organosulfur com pounds, the desulfurization of DBT at the beginning of the reaction became relatively slower until day 3 (72 h), but af ter that the degradation pattern was faster and almost the same as that of a single compound. Finally, DBT can be desulfurized up to 86.00% on day 5 (120 h). While the desulfurization of the compound 4,6dibutyl DBT from the beginning to the end of incubation was very slow, it could only be desulfurized as much as 7% on day 5. Figure 4 shows the mixture of three compounds, with DBT desulfurization occurring at the reaction commencement and becoming relatively slower until the 3 rd d (72 h). Sub sequently, the desulfurization pattern reached almost the same as a single or two compounds mixture, and finally arrives at 86.00% on day 5 (120 h). In cases where DBT, 4,6dibutyl DBT, and 4hexyl DBT are biodesulfurized together, the T7b strain growth patterns observed are slightly different, and a relatively slow growth is exhibited for 2 d, and this only reaches OD660 1.79 after 4 d of incubation. Figures 2, 3, and 4 show that this desulfurization activity to be a little lower compared to DBT desulfurization with a single compound (OD 660 2.17).

Time course of a mixture of DBT, 4,6dibutyl DBT, and 4hexyl DBT
The desulfurization pattern was similar, despite the slight decline in DBT desulfurization over a 3 d period. For instance, in a case where residue from single DBT desulfurization is 61.36 mg/L or 38.64% DBT is desulfu rize, a DBT and 4,6dibutyl DBT mixture have a slightly higher residue of 64.0 mg/L (36.00% desulfurized). Figure  4 shows that despite a slight reduction in DBT desulfuriza tion, the T7b strain was able to desulfurize 4hexyl DBT compounds more effectively (41%) for 120 d. Meanwhile, the desulfurization rate of 4,6dibutyl DBT in a mixture of three compounds was similar to the desulfurization rate of this compound in a mixture of two compounds was 6.66%, observed after 120 h incubation at 27°C with shaking at 273 rpm.

Biodesulfurization of light gas oil
The T7b strain was also tested for LGO organosulfur com pounds desulfurization capability through hydrodesulfu rization (HDS) treatment with 280 mg/L sulfur content.
LGO contains various aromatic sulfur compounds, in cluding DBT, BT, and derivatives. Also, the T7b strain FIGURE 4 Biodesulfurization of a DBT, 4-hexyl DBT, and 4,6dibutyl DBT mixture by growing S. subarctica T7b cells in MSSF-TD medium. The data above represents the duplicate sample analyses mean results and the average relative standard deviation for all data points did not surpass 5%.
reduced the total sulfur content of LGO by 116 mg/L (41.00%), leaving a residue of 164 mg/L (59.00%) within 5.5 d of incubation in the biphasic batch culture.

FIGURE 5
LGO GC-AED pattern for the degradation of an organosulfur compound mixture in light gas oil by growing S. subarctica T7b cells. A. Commercially hydrodesulfurized LGO (S=280 mg/L). B. After desulfurization by S. subarctica T7b, (S=164 mg/L). The strain was cultivated aerobically in an MSSF-LGO medium (5 mL MSSF medium containing 1 mL of LGO) at 27°C with shaking at 273 rpm for 132 h (5.5 d).

Discussion
This study describes the biodesulfurization and HDS prop erties exhibited by S. subarctica T7b on DBT and its derivatives, as well as LGO. The strain was discovered to effectively desulfurize DBT alone or in the presence of 4,6 dibutyl DBT, a mixture of DBT, 4,6dibutyl DBT, and 4 hexyl DBT, desulfurize an HDStreated LGO containing a complex mixture of organosulfur compounds. According to Kobayashi et al. (2001), the alkyl DBTs relative desulfu rization activities were reduced compared to all the alkyl substituentgroup carbon atoms. However, each one with a six carbonalkyl substituent group was not desulfurized. Figure 2 showed that 100 mg/L (0.54 mM) DBT was desulfurized within 120 h. Conversely, T7b strain was un able to utilize DBT as the sole carbon source. DBT, 4,6 dibutyl DBT, and 4hexyl DBT were simultaneously re moved by S. subarctica T7b cells as the sole sulfur source. Chen et al. (2008) reported the simultaneous desulfuriza tion of DBT and 4,6dimethyl DBT mixture by Mycobac terium sp. ZD19 tends to occur at a lower activity level compared to cases where the process is carried out sep arately. This shows that inhibition occurs in the mixture due to substrate competition (Chen et al. 2008). Further more, the desulfurization activity was discovered to be in fluenced by alkyl substituent group types and positions (Kobayashi et al. 2001). Tang et al. (2012) reported that degradation of DBT in the presence of thiophene tends to be influenced by Pseudomonas delafieldii R8 cells.
These bacteria are able to grow and desulfurize various organosulfur compounds contained in LGO. The results show that the T7b strain is a potential bacterium for LGO aromatic sulfur compounds desulfurization. However, the microorganism's biodegradability is lower in light gas oil compared to the degradation of a single compound or mix tures containing only a few aromatic sulfur compounds types. The apparent substrate competitive inhibition re duced the desulfurization activity exhibited by each alkyl DBT mixture. Furthermore, Zhang et al. (2013) reported that DBT biodesulfurization to be influenced by activity level and alkyl DBT presence in the reaction mixture. This phenomenon was also observed in LGO biodesulfurization due to the alkylDBT and BT complex mixtures present (Kobayashi et al. 2001). In the shorter cultivation time, the LGO aromatic sulfur T7b desulfurization percentage was discovered to be 41% after incubation for 5.5 d. This result is lower compared to the 71.42% attained after culturing for 7 d with P. agglomerans D23W3, as reported by Bhatia and Bhatia and Sharma (2010). In addition, Rhodococcus sp. IMPS02 is able to remove around 60% of the crude oil's organosulfur from crude oil after incubation at 30°C for 7 d (Castorena et al. 2002). The MG1 consortium and Rhodococcus strain IGTS8 reduced the sulfur content of diesel oil by 25% for 7 d of incubation (Awadh et al. 2020).

Conclusions
T7b strain cells successfully desulfurized and converted 88.29% of DBT to a 2HBP metabolite. Meanwhile, for DBT and 4,6dibutyl DBT combination, the desulfuriza tion percentages were 86.40% and 7.00%, respectively and the DBT, 4hexyl DBT, and 4,6dibutyl DBT mixture were respectively desulfurized in 84.40%, 41.00%, and 6.66% percentages for 5 d incubation. Furthermore, the desulfur ization rate of a DBT compound is slightly higher com pared to when mixed with other aromatic organosulfur compounds. T7b strain also desulfurized 41% of aromatic sulfur compounds in LGO for 5.5 d at 27°C with recipro cal agitation at 273 rpm.