April 2019

Special Focus: Clean Fuels

Ultra-deep desulfurization of FCC gasoline: Reactive adsorption strategy

With an increase in concern for the environment, countries around the world are implementing stringent environmental regulations and are continuously pressuring refiners to reduce the amount of sulfur in transportation fuels—primarily gasoline and diesel—to combat airborne pollution caused by vehicles.

Tao, Y., SINOPEC Engineering Inc.

With an increase in concern for the environment, countries around the world are implementing stringent environmental regulations and are continuously pressuring refiners to reduce the amount of sulfur in transportation fuels—primarily gasoline and diesel—to combat airborne pollution caused by vehicles. In China, the sulfur concentration in gasoline has been regulated to less than 10 parts per million (ppm) since 2017. Commercial gasoline is mixed by different cuts from processes such as fluid catalytic cracking (FCC), reforming, isomerization, and alkylation. Generally, due to the pre-hydrotreated feedstock for reforming, isomerization and alkylation units, gasoline from these units contains little or no sulfur. Significant amounts of sulfur exist in the gasoline from FCCUs that use atmospheric residues or vacuum distillates as the feedstock. Consequently, FCC gasoline becomes the most significant sulfur contributor. In China, FCC naphtha contributes to approximately 75% of the gasoline pool, as well as about 90% of sulfur.

According to OPEC’s 2018 World Oil Outlook, new conversion unit capacity is projected to reach 3.3 MMbpd by 2023—FCC/residue FCC (RFCC) represents 29% of new conversion capacity (nearly 1 MMbpd) within the forecasted time frame.1 Due to the paraffinic nature of domestic feedstocks, most of these new units will be in Asia and the Middle East. Accordingly, 6.7 MMbpd of new desulfurization capacity is projected within the forecasted time frame. Of these additions, 1.7 MMbpd is for naphtha processing, with 900,000 bpd for gasoline. The gasoline additions relate primarily to the processing of FCC naphtha to ultra-low-sulfur standards. Most of the new desulfurization capacity is forecast to be built in the Asia-Pacific and Middle East regions.

Sulfur species in FCC naphtha

Although sulfur impurities may differ from the feedstock for FCC units, the main sulfur components in FCC gasoline are mercaptans, sulfides, thiophenes and alkylthiophenes, tetrahydrothiophenes, thiophenols and benzothiophenes. A semiquantitative distribution of sulfur species in various feedstock found that the main sulfur compounds are thiophenes, which may represent more than 60 wt% of the total sulfur content in FCC gasoline.2 Numerous works indicate that thiophenic compounds (including thiophene, benzothiophene and their alkylated derivatives) are much less reactive than mercaptans and sulfides during the hydrodesulfurization (HDS) of naphtha. This requires a robust HDS process, which makes HDS a very costly option for deep desulfurization.


Ultra-deep desulfurization technologies for FCC gasoline

The challenge for ultra-deep desulfurization of gasoline is to efficiently remove thiophenic compounds of FCC naphtha while simultaneously preventing a loss in octane number and a high yield of naphtha. Several commercially running technologies could be obtained, as shown in TABLE 1. From the principle of desulfurization, two distinguishable strategies have been commercially developed: HDS and reactive adsorption desulfurization (RADS).

Fig. 1. Reaction scheme of HDS and RADS.
Fig. 1. Reaction scheme of HDS and RADS.

HDS is traditionally used in refineries to reduce the sulfur content in fuels. Typically, the HDS process involves cobalt molybdenum aluminum oxide (CoMo/Al2O3) or nickel molybdenum (NiMo)/Al2O3 catalysts with hydrogen (H2) to convert the various sulfur compounds to hydrogen sulfide (H2S) and sulfur-free organic compounds at relatively severe operating conditions (FIG. 1), commonly at temperatures in the range of 300°C–450°C and at H2 pressures of 3 MPa–5 MPa, while the resulting H2S is eventually converted into elemental sulfur by the Claus process.

The main reaction in catalytic cracking is the transformation of long chain paraffins via β-scission of intermediate carbenium ions; thus, catalytic cracking will produce equal amounts of olefins and paraffins. Under mild HDS conditions, H2S can react with olefins in the reactor, resulting in the formation of recombinant mercaptans that are linear or branched thiols of 5–12 carbons. This causes sulfur to be retained in the final product, limiting the effectiveness of the HDS process.1 Moreover, the large concentration (20 wt%–40 wt%) of compounds with aromatic rings or olefins in FCC naphtha makes HDS costly for ultra-deep desulfurization, since the non-negligible loss of octane number happens due to olefin saturation during HDS. Hence, as detailed in TABLE 1, efforts to preserve octane number have been made by two strategies: selective HDS minimizing olefins and deep desulfurization associated to octane recovery through alkane isomerization.

FCC naphtha could be separated into three main fractions:

  1. Light fraction with many olefins but few mercaptans, as sulfur impurities can be withdrawn through caustic washing
  2. Medium fraction with a low octane number—essentially thiophene and light alkylthiophene
  3. Heavy fraction with a relatively low concentration of olefins but most of the sulfur impurities.

For several modified HDS units, the etherification process for light fractions is usually introduced to obtain ether’s octane booster (e.g., tert-amyl methyl ether) for compensation of octane loss. However, as part of a clean fuels program, an ethanol blending program has become another regulation in places such as China and India. This has resulted in a significant decrease in the demand for ester’s octane booster; thus, the combination of the etherification process and HDS becomes less competitive and attractive for refiners.


Reactive adsorption desulfurization normally refers to a process that uses a metal-based solid sorbent to capture sulfur from sulfur compounds to form metal sulfide and a sulfur-free organic molecular (FIG. 1). A commercially proven sulfur removal technology,a based on the principle of RADS, has been developed. This process is designed to remove sulfur from full-range naphtha—from 2,000 μg/g feed sulfur to less than 10 μg/g of the product’s sulfur—in a one-step process by the adsorption of sorbent in the presence of hydrogen. The author’s company purchased this technology in 2007 and has continuously improved the process.

Sorbent’s selectivity and capacity for sulfur are keys to RADS performance. Many researchers dedicated to the development of RADS sorbents have found that the ideal sorbent is nickel/zinc oxide (Ni/ZnO) based. A series of commercial metal oxides such as Al2O3-, silicon dioxide (SiO2-), titanium dioxide (TiO2-), zirconium dioxide (ZrO2-) and Y zeolite can be used. Considering the cost of sorbent for industrial operations, kneading or precipitation mixing of ZiO, NiO, pseudoboehmite and SiO2 is normally applied (TABLE 2). The sorbent may consist of silica (20 wt%–60 wt%), alumina (5 wt%–15 wt%), ZiO (15 wt%–60 wt%) and Ni and/or Co (a few wt%). Although Ni/ZnO is thought to be the active component, literature has confirmed that the presence of SiO2 and Al2O3 could improve the RADS performance of FCC gasoline and reduce RON loss. This was discovered while investigating RADS experiments of FCC gasoline over a Ni/ZnO-SiO2-Al2O3 adsorbent.3 A performance test4 found that high porosity and uniformly dispersed metal components of adsorbent are keys to high RADS activity. Small-sized ZnO particles can exhibit high RADS activity, as well.5

Mechanism of RADS

Fig. 2. Proposed mechanism of RADS via Ni-ZnO-based sorbent during reaction and regeneration. Process order: (1) adsorption of sulfur, (2) transfer sulfur atom from Ni to ZnO, (3) adsorption of sulfur, (4) oxidation of sorbent during regeneration, and (5) reduction of sorbent during regeneration.
Fig. 2. Proposed mechanism of RADS via Ni-ZnO-based sorbent during reaction and regeneration. Process order: (1) adsorption of sulfur, (2) transfer sulfur atom from Ni to ZnO, (3) adsorption of sulfur, (4) oxidation of sorbent during regeneration, and (5) reduction of sorbent during regeneration.

The RADS mechanism over a Ni/ZnO-based sorbent has already been proposed by many works. As indicated in FIG. 2, it is generally accepted6 that the RADS process, in the presence of H2, may happen via three steps:

  1. Nickel sulfide (NiS) is formed when Ni-active sites are exposed to organosulfur compounds.
  2. NiS reacts with H2 to restore Ni-active sites with the release of H2S.
  3. The resulting H2S reacts with ZnO to form zinc sulfide (ZnS).

An investigation on the reaction kinetics of thiophene over Ni/ZnO adsorbent by thermal gravimetric analysis identified a rapid sulfur chemisorption on Ni followed by a nucleation-controlled conversion of ZnO to ZnS.7 For the first step, thiophene decomposition was thought to be the rate-limiting step, where a highly dispersed Ni metal was favorable—while, in the final step, thiophene diffusion determined the rate of ZnO sulfidation to ZnS. However, an in-situ X-ray absorption spectroscopy study7 suggested the formation of nickel sulfide (Ni3S2) instead of NiS. The study further revealed the important role of hydrogen in RADS by comparing RADS performance in nitrogen and hydrogen atmospheres. Hydrogen facilitates the decomposition of sulfur compounds on the active Ni species, the formation of Ni3S2 and the transformation of sulfur from Ni3S2 to ZnO. Metallic Ni, as the active nickel species, is preserved until most of the ZnO is converted into ZnS. RADS researchers are working to obtain more significant detail on their efforts.

RADS process

FIG. 3. Simplified scheme of a commercially running RADS process.
FIG. 3. Simplified scheme of a commercially running RADS process.

A simplified scheme of the RADS process is shown in FIG. 3. The FCC gasoline feed—mixed with H2—is heated to a designed temperature and vaporized. It then enters the bottom of a desulfurization reactor, where adsorption desulfurization reaction takes place in the fluidized bed of sorbent. The process operates in a gas phase to avoid the retention of the oil in the porosity. H2 is necessary to limit coking and to help the extraction of sulfur from the organic molecules, and it is captured by the sorbent.

After the reaction, the cleaned effluent leaves the top of the reactor. The spent sorbent carrying the adsorbed sulfur from the vapor is transported to a lock hopper via a sorbent receiver. In the process of sorbent regeneration, oxidation reactions occur in a fluidized bed among air and the spent sorbents; the regenerated sorbent is lifted by nitrogen to a regenerator receiver and then to the lock hopper. The sorbent recycles between the reactor and the regenerator, and is controlled by the lock hopper to provide the corresponding sorbent flow rate required by the adsorption reaction.

TABLE 3 provides the typical conditions and performance of a commercially running RADS process. Under the optimized industrial conditions, sulfur content of the final product can be lower than 10 ppm, and octane number loss can be limited to less than 0.5, while the liquid yield can be more than 99.5%. These results demonstrated that RADS can economically reduce the sulfur content of gasoline to less than 10 ppm, with minimal octane loss and near-zero volume loss.

Advantages of RADS

The RADS process has several advantages over HDS. RADS has no problem with regard to H2S inhibition and mercaptan recombination, since no free H2S is generated. The sulfur atom is retained on the sorbent, while the hydrocarbon part is released back into the stream. H2S is not released back into the product stream, and, therefore, it decreases the effluent sulfur concentration. Furthermore, the unique selectivity for sulfur removal at low levels of olefin hydrogenation minimizes octane loss. The sorbent also promotes olefin hydroisomerization, which helps counter octane loss from olefin hydrogenation.

RADS requires lower H2 feed purity; therefore, H2 from the reformer (more than 70 v%) is an acceptable source. It also requires less energy consumption (as low as 4 KgEO/t·feed), since there is no pre-splitting of the FCC feed stream and full-range naphtha is applicable. Regenerated sorbent, with sustained stable activity, allows synchronization of maintenance schedules with the FCC unit.

Another interesting issue of RADS is sulfur reactivity. Relative reactivity is normalized to thiophene (FIG. 4).8 Thiophene is one of the hardest sulfur compounds to remove for either HDS or RADS. However, there is an enhancement in activity for the other sulfur species for the RADS sorbent; consequently, RADS offers high flexibility regarding gasoline endpoint changes, and the operational upsets that the feed sulfur speciation significantly varies.


Clean fuels (e.g., ultra-low sulfur content) are the key to low emission rates and clean air. However, gasoline desulfurization, with minimal or no loss of octane number, is a challenge. Either the HDS or RADS approach can obtain ultra-low-sulfur-content gasoline; however, the ability from the unique sorbent to retain high-sulfur conversion, while minimizing olefin hydrogenation and the subsequent octane loss, presents a significant advantage of RADS over HDS technologies. This makes RADS an alternative and competitive choice for the refiners. HP


aRefers to Sinopec’s S-Zorb process


  1. OPEC’s World Oil Outlook, www.opec.org, 2018.
  2. Srivastava, V., “An evaluation of desulfurization technologies for sulfur removal from liquid fuels,” RSC Advances, 2012.
  3. Fan, J., et al., “Research on reactive adsorption desulfurization over Ni/ZnO−SiO2−Al2O3 adsorbent in a fixed-fluidized bed reactor,” Industrial & Engineering Chemistry Research, 2010.
  4. Ullah, R., et al., “Comparison of reactive adsorption desulfurization performance of Ni/ZnO-Al2O3 adsorbents prepared by different methods,” Energy & Fuels.
  5. Zhang, Y., et al., “Ultra-deep desulfurization via reactive adsorption on Ni/ZnO: The effect of ZnO particle size on the adsorption performance,” Applied Catalysis B Environmental, 2012.
  6. Babich, I., and J. Moulijin, “Recent advancement on deep desulfurization of gasoline and diesel oil,” Fuel, 2003.
  7. Huang, L., et al.,In situ XAS study on the mechanism of reactive adsorption desulfurization of oil product over Ni/ZnO,” Applied Catalysis B Environmental, 2011.
  8. Laan, J., “ConocoPhillips S Zorb gasoline sulfur removal technology: Unique chemistry, proven performance, and optimized design,” https://www.scribd.com/document/224392896/Szorb-Process.

The Author

Related Articles

From the Archive