Synthesis of a Key Linezolid Intermediate via Fixed-Bed Continuous-Flow Hydrogenation: Insights into Catalyst Grade Selection and Optimum Reaction Conditions for Scale-Up

Abstract

3-Fluoro-4-morpholinoaniline 1 is a key precursor in the synthesis of the high-volume antibacterial drug Linezolid and is typically obtained via catalytic hydrogenation-mediated nitro-group reduction in 4-(2-fluoro-4-nitrophenyl)morpholine 2. Ability to carry out this transformation in a safe and scalable manner, without significant deshalo impurity formation, is of paramount importance in defining an efficient process for the titular API. Herein, we describe a detailed study into fixed-bed-mediated hydrogenation of 2 in order to delineate an easily scalable means of accessing 1 in continuous flow that incorporates elements of simplicity and efficiency. The effect of catalyst (Pd/C and Raney Ni) grade in conjunction with other parameters (such as time, temperature, solvent, and concentration) has been investigated. Further, catalyst stability and longevity were evaluated by conducting a time-on-stream study. The best reaction conditions afforded (a) >99.5% product formation, (b) near quantitative isolated yield of 1, and (c) a space–time yield of 1.65 kg/L/h, as opposed to 0.0058 kg/L/h that could be obtained in conventional batch techniques.

Keywords

Linezolid - Reduction of nitro compounds - Continuous-flow hydrogenation - Fixed-bed reactor - Heterogeneous catalysis

Significance

Fixed-bed continuous-flow hydrogenation using commercially available “flow-friendly” grades of Pd/C and Raney Ni has been investigated for the synthesis of the key intermediate of the antibacterial drug, Linezolid. The study provides critical insights into the choice of catalyst grade, residence time, temperature, and solvent for ensuring complete consumption of starting material while minimizing the formation of impurities (particularly via defluorination). The delineated process is readily adaptable for large-scale production and adheres to the principles of greenness and sustainability.

Introduction

Linezolid, an oxazolidinone antibiotic developed by Pfizer and approved by the US FDA in 2000 under the trade name Zyvox, represents a major advancement in the treatment of severe Gram-positive infections. The drug demonstrates potent activity against Streptococcus pneumoniae, methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus faecium (VREfm), and remains the only FDA-approved therapeutic option for multidrug-resistant pneumonia and MRSA-associated skin infections.[1] Needless to say, the Linezolid market share is expected to reach USD 1.8 Billion by 2031, growing at a 5.3% CAGR from 2024 to 2031.[2] Given the significant production volume associated with this drug, establishing an efficient and scalable synthesis of Linezolid and its intermediates is of paramount importance.

A key transformation in the multistep synthesis of Linezolid involves the catalytic hydrogenation of the fluoro-substituted nitrobenzene 2 to afford the corresponding fluoro-substituted aniline 1 ([Scheme 1]). Several approaches employing homogeneous,[3] heterogeneous,[4] and novel custom-designed catalytic systems[5] have been reported for this transformation.

Scheme 1 Hydrogenation of 2 provides 1, the key intermediate in the synthesis of Linezolid.

In conventional batch reactors, hydrogenation of 2 is typically carried out at ≤10 bar and ≤100 °C due to safety considerations.[6] However, such pressure and temperature limitations often result in long reaction times (8–24 h).[7] Batch hydrogenation processes do in general suffer from poor mixing and mass-transfer inefficiencies, safety concerns related to the handling of pyrophoric catalysts and hydrogen gas, and cumbersome downstream processing steps such as catalyst filtration. Furthermore, the scale-up of batch hydrogenation to multi-tonnage production poses significant engineering and operational challenges, particularly in terms of reactor design, power, and utility requirements.[6] [7]

In this context, fixed-bed continuous-flow hydrogenation has emerged as an attractive alternative for gas–liquid–solid (G/L/S) multiphase reactions over the past two decades.[8] [9] The significant advantages they confer include (1) efficient mixing in reactions that are limited by mass-transfer limitations, (2) facile and safe handling of pyrophoric catalysts, (3) simplified downstream processing, and (4) capability of producing a high-throughput.[6] [7] [8] [9] Unsurprisingly, there is a growing interest in investigating the use of fixed-bed continuous-flow hydrogenation for accessing several pharmaceutically relevant intermediates such as 1. Some of the salient research carried out in the synthesis of 1 via hydrogenation of 2 in flow has been tabulated in [Table 1].

Table 1 Reported continuous-flow hydrogenation of 2 to 1.

Catalyst	Optimized conditions	Ref.
Pd/C slurry	85 °C, 7 bar, 3 min	[10]
Pd(0) coated on silicycle	130 °C, 100 psi, 20 s	[11]
Pd-coated static mixers	120 °C, 20–24 bar, 7 min	[12]

While the aforementioned studies do represent a notable progress in the application of flow chemistry for synthesizing 1 from 2, we felt that improvement in the following areas is required for the process to be viable in terms of commercial application and sustainability:

• Scalability: Identification and use of commercially available catalyst grades with suitable particle sizes.[6] [13]

• Use of benign reagents/solvents: Implementation of safer, cost-effective, and environmentally favorable media that facilitate scale-up and align with green and sustainable chemistry principles.[14]

Against this background and given our on-going endeavors in the application of flow chemistry techniques to synthesis of nutraceuticals and pharmaceuticals,[15] we directed our efforts toward developing an improved flow-based approach to the crucial Linezolid intermediate 1. At the outset, our investigations specifically focused on developing a fixed-bed continuous-flow hydrogenation of 2 that employs commercially available catalysts and would therefore be readily amenable for synthesis of 1 even at industrial scale. It is important to note at this point that not all commercially available catalysts are conducive for this purpose. Commercial grade catalysts used for batch hydrogenations typically have small particle sizes (~40 μm).[6] [13] When used in a fixed-bed reactor, such catalysts tend to generate high pressure drops, which can quickly overwhelm the pumps and stall the operation, or add additional load onto the utilities and make the process essentially unsustainable in the long run.[16] Hence, for the hydrogenation of 2, we wanted to investigate commercially available “flow-friendly” larger particle size Pd/C or Raney Ni catalysts that would produce low pressure drops easily manageable by the pumps employed.

Results and Discussion

Five different grades of commercially available Pd/C catalyst ([Table 2]) were chosen for our initial investigations into the hydrogenation of 2 in flow. The continuous-flow setup assembly employed for the NO₂ ➔ NH₂ reduction in 2 is represented in [Fig. 1].

Table 2 Pd/C (10 %, dry) catalysts screened.

No.	Grade	Particle size (μm)
1	Noblyst F1611	150–190
2	Noblyst F1612	150–190
3	Noblyst F1603	625–675
4	Noblyst F1605	625–675
5	Noblyst F1607	625–675

Fig. 1 Schematic representation of the continuous-flow setup, where a solution of 2 dissolved in suitable solvent (a) is pumped through a liquid-delivery pump (b) which is connected to a hydrogen source (c) via a mass-flow controller (MFC, d) and check-valve (e) through a gas–liquid T-mixer (f) which is further connected via a preheating loop (g) to a fixed-bed column reactor (h) that is maintained under desired conditions of temperature in a temperature-controller bath (i) and requisite pressure maintained through a fixed back-pressure regulator (j) ultimately leading to the outlet stream exiting the reactor (k) containing the solution of 1.

Effect of Residence Time and Temperature on Reaction Profile

Hydrogenation of 2 was studied at different residence times and temperatures with the five selected Pd/C catalysts ([Table 3]). For these initial experiments, THF was chosen as the solvent (based on a batch literature precedent)[17] and the concentration of 2 in the input feed to the fixed-bed set at 0.04 g/mL. Complete consumption of 2 with little to no formation of 3 — a common desfluoro process impurity observed in the manufacturing of Linezolid ([Scheme 2])[18] — was observed for all five Pd/C catalysts under the set of residence times and temperatures studied. However, formation of impurities other than 3 appeared to be uniformly affected by the combination of residence time and temperature employed. In particular, increase in residence time at a given temperature led to increased impurity formation. Among the five Pd/C grades, Noblyst F1611 consistently gave minimum impurity formation under the residence times and temperatures investigated.

Table 3 Time–temperature screening for Pd/C based hydrogenation of 2.

No.	Catalyst	Temp. (°C)	Time (min)	HPLC (%AUC)
				1	3	Σ (others)
1	F1611 [150–190 μm]	30	1	99.16	–	0.84
			3	98.65	–	1.35
		50	1	99.52	0.13	0.35
			3	99.46	0.07	0.47
		70	1	99.59	–	0.41
			3	99.23	–	0.77
2	F1612 [150–190 μm]	30	1	98.06	–	1.94
			3	96.22	–	3.78
		50	1	97.55	–	2.45
			3	96.78	–	3.22
		70	1	98.92	–	1.08
			3	98.49	–	1.51
3	F1603 [625–675 μm]	30	1	99.12	–	0.88
			3	98.01	–	1.99
		50	1	98.98	–	1.02
			3	98.32	–	1.68
		70	1	99.42	–	0.58
			3	99.04	–	0.96
4	F1605 [625–675 μm]	30	1	98.61	–	1.39
			3	98.07	–	1.93
		50	1	98.8	–	1.2
			3	97.99	–	2.01
		70	1	98.81	–	1.19
			3	98.11	–	1.89
5	F1607 [625–675 μm]	30	1	98.79	–	1.21
			3	96.96	–	3.04
		50	1	99.25	–	0.75
			3	98.97	–	1.03
		70	1	99.55	–	0.45
			3	99.03	0.11	0.86

All reactions were performed using a filtered solution of 1 g of 2 in 25.0 volumes of THF (0.04 g/mL concentration) through a fixed-bed reactor having a liquid volume of 1.75 mL (calculated based on 40% voidage), maintained under 6–7 bar pressure and a hydrogen gas flow rate of 120 NmL/min.

Scheme 2 Defluorination during hydrogenation of 2 leads to formation of 3, which in turn generates desfluoro Linezolid.

The promising results obtained with Pd/C encouraged us to assess the efficiency of NO₂ ➔ NH₂ reduction in 2 using commercially available “flow-friendly” grades of Raney nickel as well. Since Raney nickel is more economical than Pd/C, a flow-based synthesis of the high volume intermediate 1 using Raney nickel would be far more economically viable, thereby being of tremendous interest, in an industrial setting. Two grades of commercially available Raney nickel ([Table 4]) were chosen for our trials.

Table 4 Raney nickel catalysts screened.

No.	Grade	Particle size (μm)
1	Raika-G® CT-8588	150
2	TIMECAT-FJRE2	200

Investigation of different residence times and temperatures with the two selected grades of Raney nickel was conducted using THF as the solvent ([Table 5]). The concentration of 2 in the input feed to the fixed bed was set at 0.04 g/mL. Similar to Pd/C, complete consumption of 2 was observed in all cases. However, unlike Pd/C, hydrogenation of 2 with either of the two grades of Raney nickel resulted in the formation of the desfluoro impurity 3, particularly with increased residence time and temperature. Indeed, formation of 3 increased to nearly 10% at 90 °C with even 1 min residence time. Minimum impurity formation was observed with Raika-G® CT-8588 (150 μm particle size) at 1 min residence time.

Table 5 Time–temperature screening for Raney nickel based hydrogenation of 2.

No.	Catalyst	Temp. (°C)	Time (min)	HPLC (%AUC)
				1	3	Σ (others)
1	Raika-G® CT-8588 [150 μm]	30	1	99.59	0.32	0.09
			3	98.74	1.19	0.07
		50	1	99.05	0.86	0.09
			3	98.54	1.27	0.19
		70	1	98.58	1.24	0.18
			3	97.78	1.91	0.31
2	TIMECAT- FJRE2 [200 μm]	30	1	98.37	0.18	1.45
			3	98.73	0.83	0.44
		50	1	98.01	0.52	1.47
			3	99.19	0.18	0.63
		70	1	98.06	1.56	0.38
			3	96.15	3.32	0.53

All reactions were performed using a filtered solution of 1 g of 2 in 25.0 volumes of THF (0.04 g/mL concentration) through a fixed-bed reactor having a liquid volume of 1.75 mL (calculated based on 40% voidage), maintained under 6–7 bar pressure and a hydrogen gas flow rate of 120 NmL/min.

Solvent Screening

THF, DMF, DMAc, DCM, ethyl acetate, 2-methyl-THF, and 1,4-dioxane were selected for the solvent screening studies. These solvents are hydrogenation-compatible and have the ability to dissolve both 1 and 2 readily. Hydrogenation of 2 was studied at 30 °C and with 1 min residence time in the seven chosen solvents using both Pd/C (Noblyst F1611) and Raney Ni (Raika-G® CT-8588) as catalyst ([Table 6]). Unlike the experiments described above, concentration of 2 in the input feed to the fixed bed in these experiments was increased to 0.1 g/mL.

Table 6 Solvents screened for hydrogenation of 2 (%AUC provided only for 1 and unreacted 2).

No.	Catalyst	Solvent	HPLC (% AUC)
			1	2
1	Pd/C F1611 [150–190 μm]	1,4-Dioxane	99.41	–
		DMAc	95.05	0.96
		DMF	98.97	0.19
		2-Methyl-THF	99.54	–
		THF	99.62	–
		Ethyl acetate	99.5	–
		DCM	99.74	–
2	Raney Ni Raika-G® CT-8588 [150 μm]	1,4-Dioxane	44.98	53.03
		DMAc	18.68	76.02
		DMF	20.72	73.27
		2-Methyl-THF	86.51	12.37
		THF	32.6	67.14
		Ethyl acetate	99.11	–
		DCM	98.88	0.99

All reactions were performed using a filtered solution of 1 g of 2 in 10.0 volumes of each solvent (0.1 g/mL concentration) through a fixed-bed reactor having a liquid volume of 1.75 mL (calculated based on 40% voidage), maintained at 30 °C under 6–7 bar with a hydrogen gas flow rate of 120 NmL/min.

Noblyst F1611 gave complete consumption of the starting material in five of the seven chosen solvents. However, with Raika-G® CT-8588, complete or nearly so consumption of 2 was observed solely in ethyl acetate and DCM. Compared to DCM, ethyl acetate is a substantially greener and more sustainable solvent. It presents lower risks to human health, exhibits reduced environmental impact, and can be produced from bioethanol.[19] Hence, ethyl acetate was chosen for the time-on-stream (TOS) experiments with Noblyst F1611 and Raika-G® CT-8588 described below.

TOS Experiments: Scaling Up Synthesis of 1 on a Multigram Scale

TOS studies were carried out to obtain a preliminary understanding of the stability and longevity of the two chosen catalysts. Implicit in our efforts was also the expectation that the TOS studies would provide an insight into the scalability and robustness of the fixed-bed continuous-flow hydrogenation process developed.

Hydrogenation of 2 to 1 was carried out with a SS 316 column (length = 15.5 cm, ID = 6 mm, volume = 4.38 mL) filled with 1.5 g of Noblyst F1611 (150–190 μm) with a void volume of 1.75 mL (calculated based on 40% voidage). This in-house fabricated fixed-bed reactor was maintained at 30 °C and 6–7 bar pressure while a filtered solution of the starting material (concentration = 0.08 g/mL in ethyl acetate) and hydrogen gas were passed through it. The liquid and hydrogen gas flows were maintained at 1.75 mL/min and 120 NmL/min, respectively, to afford a residence time of 1 min. The output stream after attainment of steady state was collected over a period of 2.167 h as 13 aliquots (amounting to 227.5 mL in total). HPLC analysis revealed complete consumption of 2 and the presence of 1 to an extent of >99.5% in all the collected aliquots ([Fig. 2a]). Thus, 18.2 g of 2 could be processed over 1.5 g of Pd/C with no loss in catalytic activity.[20]

Fig. 2 (a) TOS studies for Pd/C (Noblyst F1611 [150–190 μm]). (b) TOS studies for Raney nickel (Raika-G® CT-8588 [150 μm])

For the TOS studies with Raika-G® CT-8588 (150 μm), the SS 316 column, earlier used for TOS studies with the Pd/C catalyst, was now filled with 5.8 g of the Raney nickel catalyst with a void volume of 1.75 mL (calculated based on 40% voidage). Similar to the experiments carried out with Pd/C, a filtered solution of the starting material (concentration = 0.08 g/mL in ethyl acetate) and hydrogen gas were passed through this in-house fabricated fixed-bed reactor under the following conditions: (a) temperature = 50 °C; (b) pressure = 6–7 bar; (c) residence time = 1 min; (d) liquid flow = 1.75 mL/min; and (e) hydrogen gas flow = 120 NmL/min. Unlike Pd/C, the TOS study with Raney nickel was run over a course of 2 days, inclusive of an overnight hold-up period, and the output stream, after attainment of steady state, was collected as 37 aliquots (amounting to 684.5 mL in total). Complete consumption of 2 and the presence of 1 to an extent of >99.5% were observed in all the fifteen representative aliquots that were submitted for HPLC analysis ([Fig. 2b]). All the 37 collected aliquots (representing an input of 54.76 g processed over 5.8 g of the catalyst) were combined and concentrated under reduced pressure to obtain 47 g (99%) of 1 as a solid with 99.9% HPLC purity.

Conclusions

To summarize, we have carried out a systematic investigation into the fixed-bed continuous-flow hydrogenation of 2 catalyzed by five Pd/C and two Raney nickel catalyst grades that are both commercially available and flow-friendly. Careful screening of residence time, temperature, and solvent led to the identification of a catalyst grade (for both Pd/C and Raney Ni) that promotes complete and rapid NO₂ ➔ NH₂ reduction in 2 under mild conditions and in a green solvent with negligible impurity formation. Noblyst F1611 and Raika-G® CT-8588 — the grades identified for Pd/C and Raney nickel, respectively — showed no loss of catalytic activity even upon extended operation during the TOS studies. Space–time yield obtained for the continuous-flow hydrogenation, demonstrated herein in an in-house fabricated fixed-bed reactor, is 1.65 kg/L/h. Indeed, this is orders of magnitude higher than that obtained for conventional batch hydrogenation of 2 (e.g., 0.0058 kg/L/h calculated for the process reported in Ref. [4c]). From the perspective of sustainability and industrial application, our process is energy efficient and easily adaptable for large-scale synthesis of 1 — the key intermediate in the production of the high-volume, life-saving antibiotic Linezolid. Investigations into further improvement of the described fixed-bed continuous-flow hydrogenation process using nickel foam are currently underway and will be communicated soon.

Contributorsʼ Statement

Manish Manohar Shinde: Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft. Karuna Veeramani: Conceptualization, Data curation, Formal analysis, Methodology, Writing - original draft. Rahulvarma Katari: Formal analysis, Investigation, Methodology. Vishnuvardhana Eda: Conceptualization, Data curation, Formal analysis, Investigation, Supervision, Writing - review & editing. Saikat Sen: Methodology, Project administration, Supervision, Writing - review & editing. Srinivas Oruganti: Conceptualization, Funding acquisition, Project administration, Supervision, Writing - review & editing.

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgment

The authors thank Evonik Industries for providing free samples of the Pd/C catalysts employed in this investigation.

• References

• 1 Fernandes GFS, Campos DL, da Silva IC. et al. ChemMedChem 2021; 16: 1268-1282
• 2 Linezolid drug market size and trends 2025–2033. Comprehensive Outlook https://www.datainsightsmarket.com/reports/linezolid-drug-1487947 (accessed Septemeber 5, 2025)
• 3a Cantillo D, Moghaddam MM, Kappe CO. J Org Chem 2013; 78: 4530-4542
• 3b Sarki N, Goyal V, Tyagi NK. et al. ChemCatChem 2021; 13: 1722-1729
• 4a Einsiedel J, Schoerner C, Gmeiner P. Tetrahedron 2003; 59: 3403-3407
• 4b Bozzoli A, Branch CL, Marshall H, Nash DJ, Porter RA. PCT application WO 2005/058317 A1 Dated: June 30 2005
• 4c Alla RM, Dubey AK, Sirigiri AK, Mareedu NKK. PCT application WO 2012/114355 A1 Dated: August 30 2012 https://patents.google.com/patent/WO2012114355A1/en (accessed September 5, 2025)
• 4d Branch CL, Nash DJ. WO2005103042A1 Dated: November 3 2005
• 4e Adjabeng G, Bifulco N, Davis-Ward RG. et al. PCT Application WO 2009/032667 A1 Dated: March 12 2009
• 4f Zhang Y, Zhang J, Wang X. et al. PCT Application WO 2014/012360 A1 Dated: January 23 2014
• 5a Singha Hazari A, Frisch ML, Wen Y, Stankovic MD, Berlinguette CP. J Am Chem Soc 2024; 146: 28153-28160
• 5b Shanmugaraj K, Bustamante TM, Torres CC, Campos CH. Catal Today 2022; 388–389: 383-393
• 6 Fernandez-Puertas E, Robinson AJ, Robinson H, Sathiyalingam S, Stubbs H, Edwards LJ. Org Process Res Dev 2020; 24: 2147-2156
• 7 Xiaonan D, Xuepeng W, Xingkun C, Jisong Z. Org Process Res Dev 2021; 25: 2100-2109
• 8a Lebl R, Bachmann S, Tosatti P. et al. Org Process Res Dev 2021; 25: 1988-1995
• 8b Hoogenraad M, van der JB, Smith AA. et al. Org Process Res Dev 2004; 8: 469-476
• 8c Duan X, Yin J, Feng A. et al. J Flow Chem 2022; 12: 121-129
• 8d Masson E, Maciejewski EM, Wheelhouse KMP, Edwards LJ. Org Process Res Dev 2022; 26: 2190-2223
• 9a Yoswathananont N, Nitta K, Nishiuchi Y, Sato M. Chem Commun 2005; 40-42
• 9b Irfan M, Glasnov TN, Kappe CO. ChemSusChem 2011; 4: 300-316
• 10 Skilton R, Guthrie D, Moses R. Pd/C Slurry Transfer Hydrogenation in Continuous Flow. https://www.vapourtec.com/wp-content/uploads/2015/08/PdC-Slurry-Transfer-Hydrogenation-in-Continuous-Flow.pdf
• 11 Russell MG, Jamison TF. Angew Chem Int Ed 2019; 58: 7678-7681
• 12 Gardiner J, Nguyen X, Genet C, Horne MD, Hornung CH, Tsanaktsidis J. Org Process Res Dev 2018; 22: 1448-1452
• 13 Carangio A, Edwards LJ, Fernandez-Puertas E. et al. Process Res Dev 2020; 24: 1909-1915
• 14 Loos P, Alex H, Hassfeld J. et al. Org Process Res Dev 2016; 20: 452-464
• 15a Shukla CA, Udaykumar B, Saisivanarayana Y. et al. J Flow Chem 2022; 12: 1-7
• 15b Veeramani K, Shinde M, Eda VVR. et al. Tetrahedron Lett 2023; 116: 154344
• 15c Veeramani K, Shinde M, Eda VVR. et al. J Flow Chem 2024; 14: 481-489
• 15d Patel K, Kapdi A, Shinde MM, Veeramani K, Oruganti S. Sus Circ Now 2024; 1: a22430268
• 16 Baramov T, Hassfeld J, Gottfried M. et al. Eur J Org Chem 2017; 3921-3928
• 17a Zhang Y, Zhang J. et al. PCT Application WO 2014/012360 A1 Dated: January 23 2014
• 17b Nathalie T, James L, Karen R. et al. PCT Application WO 2005/047288 A1 Dated: May 26 2005
• 18a Linezolid (Linezolidum), Draft proposal for inclusion for The International Pharmacopoeia, WHO working document QAS/20.841/Rev1, Dated May 2022 https://cdn.who.int/media/docs/default-source/medicines/norms-and-standards/current-projects/qas20_841_linezolid.pdf (last accessed October 31, 2025)
• 18b Braude V, Finkelstein N. PCT Application WO 2007/064818 A1 Dated: June 7 2007
• 19a Methylene Chloride (DCM) Replacements from the Green Chemistry Teaching and Learning Community (GCTLC) https://gctlc.org/methylene-chloride-dcm-replacements (last accessed January 16, 2026) and resources provided therein
• 19b Renewable Ethyl Acetate from Green Chemistry for Sustainability https://chemistryforsustainability.org/safer-alternatives/renewable-ethyl-acetate (last accessed: January 16, 2026)
• 19c Prat D, Wells A, Hayler J. et al. J Green Chem 2016; 18: 288-296
• 19d Welton T. Proc A 2015; 471: 20150502
• 20 See Supporting Information for the results of an extended TOS experiment (operation time under steady state = 17 h) carried out with Pd/C [Noblyst F1611 (150–190 μm)]

Publication History

Received: 31 October 2025

Accepted after revision: 20 January 2026

Article published online:
09 February 2026