Fluorine-containing hydrocarbons are commonly used as refrigerants, aerosol sprays, and foaming agents^[1-3] 1, 1, 1, 3, 3-pentafluoropropane (HFC-245fa) as the third generation foaming agent with an atmospheric ozone depletion potential (ODP) of 0, the greenhouse effect potential (GWP) of 1030, faces the reality of being restricted and abandoned^[4-6]. Non- flammabletrans-1-chloro-3, 3, 3-trifluoropropene(HCFO- 1233zd (E))with low toxicity under normal conditionsis the fourth generation of blowing agents developed in recent years^[7-10]. Its ODP and GWP are 0.000 24 and 7.0 respectively.

More particularly, it is a monomer for synthesizing polymeric materials, and a building block for making other fluorinated compounds. Compared with cis-1-chloro-3, 3, 3-trifluoropropene (HCFC-1233zd (Z)), HCFO-1233zd (E) has higher thermos-dynamic stabi- lity, and 10% of HCFO-1233zd (E) can be isome- rized to HCFC-1233zd (Z) at 300 ℃ using fluorinated Cr₂O₃ catalysts^[11].

HCFO-1233zd(E) was synthesized in liquid phase fluorination of 1, 1, 1, 3, 3-pentachloropropane (HCC-240fa) with HF using Lewis acid catalysts^[12-17]. However, the target product was synthesized in batches, and the reaction lead to more industrial wast and serious equipment corrosion. Special attention has been paid to vapor fluorination for the synthesis of HCFO-1233zd (E) in the chemical industry, because of its potential of industrial continuous production. Until now, only several patents reported vapor fluorination routes to synthsize HCFO-1233zd (E)^[18-22]. Catalysts such as Cr₂O₃^[18], CrCl₃/Al₂O₃^[19], AlF₃^[20-21] and Cr-Ni/AlF₃^[22] are used to reach the conversion of HCC-240fa of 100%, and the selectivity of HCFO-1233zd (E), and catalyst lifespan are not ideal. Merkel et al introduced air into the reactants with the O₂/HCC-240fa molar ratio of 0.032:1, the regeneration and activation of the catalyst would be completed, and at least one more lifespan could be extended^[23]. The high-valent metal halide-supported catalysts are reported to avoid rapid carbonation on the catalyst surface caused by high temperature and the formation of deep fluorinated by-products^[24-26]. Though the yield of the process reached industrial application requirements, this patented technology does not address the stability of the catalyst. In addition, the usage of precious metals such as Sb, Ta, Mo and Nb etc limited the application of this technology.

Though satisfactory selectivity of HCFO-1233zd (E) through vapor fluorination of HCC-240fa with HFhas been obtained using Cr₂O₃ fluorination catalyst at 270 ℃ or higher, coke on the catalyst surface accele- rates the catalyst deactivation^{[18, 27-29]}.We reported highly selective synthetic route to HCFO-1233zd (E) by vapor fluorination of 1, 1, 3, 3-tetrachloropropene (HCC-1230za) with HF over the Cr₂O₃-based catalysts. HCC-1230za was synthesized through highly selective dehydrochlorination of HCC-240fa using activated carbon. 99.4% HCC-1230za conversion and 98.2% HCFO-1233zd (E) selectivity are obtained at 200 ℃, and the catalyst achieved long lifespan in this reaction.

1 Experimental details 1.1 Catalyst Preparation

The M (M = Zn²⁺, Co²⁺, Al³⁺) modified Cr₂O₃ catalysts were prepared by a precipitation method. A detailed preparation process of the catalyst Zn/Al/Cr₂O₃ is as follows: 6.815 g (0.05 mol) of zinc chloride, 24.15 g (0.1 mol) of aluminum trichloride hexahydrate, and 133.2 g (0.5 mol) of chromium trichloride hexahydrate, Zn²⁺:Al³⁺:Cr³⁺=0.1:0.2:1 (mol/mol/mol), dissolved in 270 mL of distilled water. Subsequently, 76 g of NaOH (19 mol/L) was added to a solution of Zn²⁺, Al³⁺ and Cr³⁺, and magnetically stirred until precipitation occurred and allowed to stand for 24 hours. The precipitate after aging was filtered and washed with about 2000 mL of distilled water until the pH of the solution is 7~8, then dried at 60 ℃ for 48 h, followed by a calcination at 460 ℃ for 4 h in H₂, and the oxide after calcination was 54.0 g. 3% (Percent weight) of colloidal graphite were added to the calcined oxide, and grounded by a high-speed universal pulverizer. 5%(Percent weight) of methyl cellulose solution was added to the mixture, extruded and formed into cylindrical the final catalysts, and air-dried at room temperature. Preparation process of the catalyst Co/Al/Cr₂O₃ (Co²⁺:Al³⁺:Cr³⁺= 0.2:0.1:1 mol/mol/mol) and Zn/Cr₂O₃(Zn²⁺:Cr³⁺ =1:36 mol/mol) are basically the same as above.

Zn²⁺-doped Cr₂O₃ supported over spherical γ-alumina catalyst 4% Zn/2% Cr/γ-Al₂O₃(Percent weight) was prepared as follows: 4.880 g of zinc chloride and 5.930 g of chromium trichloride hexahydrate were dissolved in 58 mL of distilled water, which is used as an immersion liquid; 50 g spherical γ-Al₂O₃ carrier was added to the impregnating solution, followed by evaporating any residual water, then dried at 70 ℃ for 72 h. Spherical γ-Al₂O₃ carrier: Φ5.0 mm, specific surface area ≥ 300 m²/g, pore volume ≥ 0.40 mL/g, bulk density 0.73 ± 0.03 g/mL. Using the same method as above to prepare catalyst 8% Zn/4% Cr/γ-Al₂O₃(Percent weight).

1.2 Catalytic fluorination

HCC-240fa is synthesized by any means well known in the literature^[30-32], and the synthesis of HCC-1230za (Scheme 1) is carried out using HCC-240fa as startng material as follow: 160 g of activated carbon is added to the quartz glass tube (1.5 m × 30 mm), using the method of upper feed and lower discharge. In order to ensure the smooth discharge of the product, a nitrogen purge is introduced during the reaction. The flow rate of nitrogen gas is 10 mL/min, the reaction temperature is controlled at 190~200 ℃, and the product is collected by ice salt bath cooling. During the whole reaction stage, GC analysis using a gas chromatograph (Agilent GC-7890) equipped with a flame ionization detector (FID) and a HP-5 (30 m) capillary column follows the reaction. A total of 3.0 kg HCC-240fa was input, the feed rate was 30 mL/h, and the unreacted HCC-240fa adsorbed by activated carbon was 0.32 kg. The theoretical product should be 2.23 kg, while the actual collected product HCC-1230za was 2.0 kg. The yield is 89.6%, and GC content of HCC-1230za is >94%.

Activity test of catalyst Zn/Cr₂O₃ is as follows: 10~20 g of the catalyst was filled into 304 stainless steel reaction tube (50 cm×20 mm), heated by tube furnace at 150~300 ℃. Due to the high corrosiveness of HF, the gas tightness of the reaction system needs to be checked during the experiment to prevent HF lea- kage. Flow rate of HF pre-heated in a chamber at 45 ℃ was carefully controlled at 60 mL/min. Using a sevenstar mass flowmeter. During the activation process, a large amount of water was generated and continuously activated for 10 h. The activation was complete until no water is formed in the reaction tube. Subsequently, HCC-1230za feed was regulated at room temperature with a liquid pump, and the feed amount of HCC-1230za was 0.6 mL/min. The molar ratio of HCC-1230za to HF was detected by acid-base titration, and the ratio was fixed at 1:10. The fluorination reaction was under normal pressure at 200 ℃. The effluent flow from the reactor was washed with NaOH solution in a scrubber for the removal of HCl and HF, the stream was further dried using NaOH pellets, and then the gaseous products were condensed to obtain a liquid sample for analysis. GC-MS (Thermo Scientific ITQ 700) was applied for identity of the organic compounds formed during the reaction. The reaction products were analyzed by before mentioned gas chromatograph. The relative composition of the products is based on peak areas, therefore do not represent the absolute yields, because of difference in response factors. When the GC content of HCFO-1233zd (E) decreased by more than 10%, the catalyst was considered to be deactivated. Activity test of catalyst Zn/Al/Cr₂O₃, Co/Al/Cr₂O₃, 4% Zn/2% Cr/γ-Al₂O₃, and 8% Zn/4%Cr/γ-Al₂O₃ is basically the same as above.

We selected consecutive three hours as the time node for material balance calculation: the total feed volume was 164.0 g, and the two-stage cooling device collected a total of 98.8 g of crude products and a total input of 164.0 g. The theoretical product yield was 120.3 g, the reaction yield was 82.1%. The crude GC analysis results obtained by the two-stage cooling device contained 0.62% HCC-1230za, 91.2% HCFC-1233zd (E), 1.7% HCC-245fa, and 4.1% difluorodichloropropylene. The material was lost during the reaction. The main reason was that high purity nitrogen was used in the entire experiment to dilute HF concentration to reduce safety risks, and some crude products were blown away by high purity nitrogen.

1.3 Characterization

The catalysts was subjected to crystal phase analysis through PANalytical X'Pert PRO Polycrystalline Powder X-ray diffractometer (Cu Kα, λ = 0.154 18 nm) (PANalytical BV, Almelo, Netherlands) with an angle reproducibility of ±0.0001°, the diffraction data of 2θ in 8°~80° was collected by θ/θ scanning method with a scanning speed of 2°/min. Chemical compositions on the surface of samples were analyzed using an X-ray photoelectron spectrometer (XPS) (Thermo Fisher Scientific ESCALAB 250Xi) equipped with an Al monochromatic X-ray source (Al Kα = 1486.6 eV, C 1s was corrected to 284.8 eV) under room temperature in high vacuum (about 1 × 10^-9 Pa). The N₂ adsorption-desorption characterization of the samples was performed on a specific surface area analyzer Micromeritics ASAP2020 V3.04 H (Micromeritics Instrument Co., Ltd., Norcross, USA) at liquid nitrogen temperature (-196 ℃). The specific surface area of the catalyst was determined using the Brunauer-Emmet-Teller (BET) method. The pore size distribution was analyzed by the adsorption-desorption isotherm through the Barret-Joyner-Halender (BJH) method using a columnar pore model.

The Brønsted and Lewis acid site of the sample were determined by V70 pyridine adsorption infrared spectrometer (Bruker, Germany) with a resolution of 4 cm^-1 and an uptake range of 400~4000 cm^-1. The catalyst sample was pressed into a thin sheet, placed in a quartz infrared cell, heated to 400 ℃ and maintained a constant temperature for evacuation pretreatment, then was degassed and dehydrated under high vacuum (1×10^-3 Pa) for 2 h, and naturally cooled to 150 ℃. The anhydrous pyridine was adsorbed at 150 ℃ for 30 min, then heated to 400 ℃ to start desorption, 400 ℃ for 0.5 h, 150 ℃ for 0.5 h, and then the Py-IR spectrum was recorded on an infrared spectrometer. The temperature-programmed desorption of ammonica (NH₃-TPD) measurement was carried out on a fully automatic multifunctional dynamic adsorber instrument DAS-7200 (Huasi Instrument Co., Ltd, China) for comparing the acidity strength of the composite catalysts. A sample of 100 mg on the quartz glass tube (i. d. = 5 mm) was first pretreated in pure N₂ from 100 to 450 ℃ at a rate of 10 ℃/ min, and kept at 450 ℃ for 2 h. followed by cooling to 100℃ in a flow of N₂ (30 mL/min). NH₃-N₂ mixture (10% NH₃, 30 mL/min) was introduced to the reactor for 2 h. to allow the complete adsorption of NH₃. Prior to measurement, the sample was purged with pure N₂ (30 mL/min) to remove physically absorbed NH₃ for 60 min. Then the sample was heated in the N₂ flow (30 mL/min) from 100 to 850 ℃ with a heating rate of 10 ℃/min., and the desorption of NH₃ was recorded with a thermal conductivity detector.

2 Results and discussion 2.1 Effect of HF/HCC-1230za molar ratio and reaction temperature

It is known that HCFO-1233zd (E) was synthesized by conducting a fluorination reaction of HF and HCC-240fa in the presence of catalyst Cr₂O₃ at 270~300 ℃ with the molar ratio of HF/HCC-240fa varied from 20 to 29. Only 82% selectivity to HCFO-1233zd (E) is obtained with HF/HCC-240fa molar ratio of 20:1, at 270 ℃, and at atmospheric pressure^[18]. In our work, the reactions of HCC-1230za with anhydrous HF were started first using Zn/Cr₂O₃ catalyst at atmospheric pressure and at 250 ℃. We investigated the effects of molar ratio of HF/HCC-1230za on the selectivity to HCFO-1233zd (E). Table 1 shows that the data of product distribution at different molar ratio of HF/HCC-1230za. The conversion of HCC-1230fa increased significantly from 91.0% to 98.8%, and the selectivity to HCFO-1233zd (E) increased from 55.2% to 81.5% as the molar ratio of HF/HCC-1230za varied from 6 to 10. At the same time, the selectivity to HCFO-1233zd (Z) and HCFO-1232zc decreased drastically from 14.2% to 0.88% and 11.0% to 6.50% respectively. It suggested that the formation of unwanted side products HCFO-1233zd (Z) and HCFO-1232zc is greatly suppressed by excess HF (Scheme 2). Even if the molar ratio of HF/HCC-1230za further increased beyond 10, little influence on the selectivity to HCFO-1233zd (E) and the amount of side products. So we choose HF/HCC-1230za molar ratio of 10:1 in the next experiment.

As shown in Table 2, the selectivity to HCFO-1233zd (E) increased remarkably with the reaction temperature varied around 200 ℃, and the selectivity to HCFO-1233zd (E) declined if reaction temperature increase. At this temperature, the conversion of HCC-1230za is also the highest. Whereas, the selectivity to HCFO-1232zc decreased from 24.5% to 5.10% as temperature increased from 150 to 300 ℃. These results indicate that the reaction of HCFO-1232zc with HF is difficult to occur in this reaction because it requires higher energy (Scheme 2). With the reaction temperature increasing from 150 to 300 ℃, HCFO-1233zd (Z) firstly decreased and then increased. This means that the reaction temperature has an effect on the transformation between two isomer HCFO-1233zd (E) and HCFO-1233zd (Z) (Scheme 2)^[11].

2.2 Effect of catalyst

Next, we observed the reactions of HCC-1230za with HF using different catalysts and an additional reaction of 10-fold excess HF at 200 ℃. From Table 3, it can be seen that the use of Zn/Cr₂O₃ catalyst resulted in 98.2% of HCFO-1233zd (E) (Table 3, entries 3). While the use of Zn/Al/Cr₂O₃, Co/Al/Cr₂O₃, 4% Zn/2% Cr/γ-Al₂O₃ and 8% Zn/4% Cr/γ-Al₂O₃ catalyst revealed 72.9%, 68.4%, 83.7% and 88.3% of HCFO-1233zd (E) respectively (Entries 1, 2, 4, and 5). The catalyst Zn/Cr₂O₃ was more effective in yiel- ding HCFO-1233zd (E) and avoiding HCFO-1233zd (Z) than other catalysts. Therefore, the addition of Zn²⁺ to Cr₂O₃ significantly increases the selectivity of HCFO-1233zd (E) (Table 3, Entry 3).

Because the stability of the catalyst is crucial in practical application^[23-26], long term reaction were conducted using the catalysts Zn/Al/Cr₂O₃, Co/Al/Cr₂O₃, and Zn/Cr₂O₃ at 200 ℃ and contact time 10 s, respectively (Table 3). The conversions remained constant (>99%) and the selectivity of HCFO-1233zd (E) gradually declined from 98.2% to 88.2% at a time on stream of 118 h using Zn/Cr₂O₃ catalyst. This indicates that Zn/Cr₂O₃ catalyst is stabled uring the 118 h testing, and the selectivity to HCFO-1233zd (E) is satisfactory (Table 3, Entry 3). Using Zn/Al/Cr₂O₃ and Co/Al/Cr₂O₃ catalysts after 112 and 120 h reaction, the conversions not significantly changed as compared with the initial of 95.2% and 94.3%, and the selectivety to HCFO-1233zd (E) were 62.9% and 58.4%(Table 3, Entries 1 and 2) respectively. Although 4% Zn/2% Cr/γ-Al₂O₃ and 8% Zn/4% Cr/γ-Al₂O₃ catalysts suffered the most severe deactivation compared with other catalysts, the results of the study demonstrate the beneficial effect of zinc loading to chromium-containing alumina on increasing the selectivity of HCFO-1233zd (E)(Table 3, Entries 4 and 5).

The generation of active CrO_xF_y species could be approached by adding Zn²⁺, Co²⁺and Al³⁺ in the Cr₂O₃-based catalysts. For example, Lee et al^[33] found that MgF₂/ Cr₂O₃ catalyst was more active than Cr₂O₃ catalyst for the synthesis of 1, 1, 1, 2-tetrafluoroethane (HFC-134a) from 2-chloro-1, 1, 1-trifluoroethane (HCFC-133a), which was related to MgCrO_xF_y species generated through the reaction of CrF₃ and MgF₂. The vacant sites of CrF₃ or CrF_x(OH)_3-x are believed to be responsible for the L acidity, whereas B acidity can be attributed to the presence of hydroxyl group. Zn/Cr₂O₃ catalyst has the suitable acidity site distribution and strength of B acid and L acid, and which give high stable and selectivity under optimum fluorination conditions.

The reactions of HCC-1230za with HF over Zn/Cr₂O₃ catalyst prepared at different calcination temperature were carried out in N₂ or H₂ atmosphere. In preparation process of the catalysts designed as Zn/Cr₂O₃-350-N₂, Zn/Cr₂O₃-350-H₂, Zn/Cr₂O₃-460-N₂, Zn/Cr₂O₃-460-H₂, Zn/Cr₂O₃-560-N₂, Zn/Cr₂O₃-560-H₂, Zn/Cr₂O₃-600-N₂, and Zn/Cr₂O₃-600-H₂, the precipitated slurry was calcined at 350, 460, 560, and 600 ℃ for 4 h in N₂ or H₂ atmosphere to obtain the final catalysts, respectively. The results are summarized in Table 4. The reaction of HCC-1230za with HF resulted in 81.9% of HCFO-1233zd (E)over catalyst Zn/Cr₂O₃-350-N₂, while 87.0% of HCFO-1233zd (E) was gained over catalyst Zn/Cr₂O₃-350-H₂ (Entry 1 vs Entry 2, Table 4). The selectivity of HCFO-1233zd (E) steadily increased from 81.9% (at calcined 350 ℃) to 98.2% (at calcined 460 ℃) (Entry 1, 2, 3, and 4, Table 4). While the selectivity decreased from 98.2% to 82.0% (entry 4, 5, 6, 7, and 8, Table 4) if calcination temperature gradually increased to 600 ℃. It can be seen that the calcination temperature has an effect on the selectivity to HCFO-1233zd (E), and the catalytic precursor calcined under H₂ was more effective than that under N₂ atmosphere (Entry 3 vs Entry 4, and Entry 5 vs Entry 6, Table 4). These results are consistent with Brunet's^[34] and Xie's ^[35]. Brunet et al ^[34] investigated the effects of activation temperature and atmosphere (N₂, H₂ or air) of Cr₂O₃ catalyst on the synthesis of HFC-134a from HCFC-133a. They concluded that Cr species of reversible oxidation state are the active sites of reaction. Xie's group demonstrated the calcination temperature have great influence on the crystalline size, surface acid sites and the molar fraction of F in the catalyst^[35]. With increasing calcination temperature, Cr₂O₃ translates from amorphous structure into crystalline phase, and the intensity of the diffraction peaks gradually becomes stronger. On the other hand, the molar fraction of F in the catalyst and the amount of surface acid sites decrease with increasing calcination temperature.

2.3 Product distribution studies

Bsed on the obtained products and by-products several plausible reaction pathways (Scheme 2) were inferred. The product HCFO-1233zd (E) can be obtained by the fluorination of HCC-1230za with HF in the presence of a suitable catalyst. Of course, this is an oversimplified reaction path that does not fully reflect the actual reactions that may occur under the established catalytic conditions. From Scheme 2, it can be speculated that the actual reaction system may be more complicated when considering the isomers of several possible intermediates. Besides the expected HCFO-1233zd (E), the other minor components formed are 1, 3-dichloro-3, 3-difluoropropene (HCFO-1232zc), and with trace amounts of 1, 3, 3-trichloro-3-fluoropropene (HCFO-1231zb), 1, 3, 3-trichloro-1, 1-difluoropropane (HCFC-242fb), 3, 3-dichloro-1, 1, 1-trifluoropropane (HCFC-243fc), 1, 1, 1, 2, 3-penta- fluoropropane (HFC-245eb), and HFC-245fa in all experiments. The selectivities to the trace amounts and other unknown products are not discussed in this work.

2.4 Characterization

of the composite catalysts XRD analyses were performed to characterize the bulk properties of Zn/Cr₂O₃ catalysts (Fig. 1). The results show that under the preparation conditions of Zn/Cr₂O₃ catalyst, the fresh catalyst (Fig. 1(1) (a)) is mainly amorphous Cr₂O₃, but it contains microcrystalline phase of Cr₂O₃ with good dispersibility, which size is estimated to be between several nanometers. The XRD pattern shows a diffraction pattern of microcrystalline phase of Cr₂O₃. The maximum peak width at half height is wider than that of single crystal Cr₂O₃. In the Fig. 1(1) (a), 2θ is a peak of graphite at 25.2°, and the 2θ of the diffraction peak of crystalline Cr₂O₃ is 44.4°, while the peaks at 34.4° and 36.2°, 63.2° and 65.0° overlap. However, no diffraction peak related to zinc species such as ZnO could be detected, suggesting the zinc species might be highly dispersed. The XRD pattern of Zn/Cr₂O₃ catalyst after 118 h reaction (Fig. 1(1) (b))is not significantly changed as compared with the fresh catalyst (Fig. 1(1) (a)). The results (Fig. 1(1) (a) and (b)) indicated that Zn/Cr₂O₃ catalysts was very stable during the fluorination process. But a new microcrystalline phase of CrF₃ (18.8°, 21.6°, and 34.6°) was found in the Fig. 1(1) (b). Although Cho et al^[36]. Found that amorphous CrO₃ with high dispersion was more advantage for the fluorination synthesis of HFC-134a from HCFC-133a than the crystalline Cr₂O₃, in our work the XRD results, as well as the evaluation results of Zn/Cr₂O₃ ca- talyst activity indicate that most of the amorphous Cr₂O₃ and finely dispersed microcrystalline phase of Cr₂O₃ together lead to high activity, high selectivity, and high stability of the catalyst. Catalyst Co/Al/Cr₂O₃ after 120 h. Reaction and the fresh one were almost amorphous Cr₂O₃(Fig. 1(2) (a) and 1(b)), which again confirmed our results.

The large surface area catalysts are desired in gas phase fluorination synthesis of HFC-134a from HCFC-133a, but the normal fluorinated metal oxides or metal fluorides have a small surface area (< 50 m²·g^-1)^[34]. BET-surface area, pore size and pore volume of the catalyst used here are summarized in Table 5. The specific surface areas are 188, 182, and 169 m²·g^-1 for the fresh Zn/Cr₂O₃, Zn/Al/Cr₂O₃, and Co/Al/Cr₂O₃ catalysts, respectively. This indicating that the specific surface area of these composite catalysts calcined at 460 ℃ in H₂ atmosphere is larger than the fluorination catalyst already reported. The specific surface area of Zn/Cr₂O₃ catalyst after 118 h reaction is 448 m²·g^-1. However, the specific surface areas are 18.2 and 66.9 m²·g^-1 for Zn/Al/Cr₂O₃ after 112 h reaction and Co/Al/Cr₂O₃ catalyst after 120 h reaction, respectively. Thus, the conversion of HCC-1230za and the selectivity to HCFO-1233zd (E)are related to the specific surface area of the Zn/Cr₂O₃ catalyst.

XPS spectra were used to identify the surface Cr species of catalyst Zn/Cr₂O₃, Co/Al/Cr₂O₃, and Zn/Al/Cr₂O₃(Fig. 2). For the fresh Zn/Cr₂O₃ catalyst, two kinds of Cr species can be attributed respectively(Fig. 2 (1)(a) )^[37]. Then, the spectrum of the Cr 2p for Zn/Cr₂O₃ catalyst after 118 h reaction, Co/Al/Cr₂O₃ catalyst after 120 h. reaction, and Zn/Al/Cr₂O₃ catalyst after 112 h. Reaction became broadening, and were shifted to higher binding energy(Fig. 2 (1) (b), (2) (a), and (2) (b)). The broad Cr 2p_3/2 peak of Zn/Cr₂O₃ sample was fitted by three components at 578.6, 579.9 and 582.7 eV, respectively. The peak at 578.6 eV is assigned to the Cr(OH)₃ phase. The peak at 579.9 eV is ascribed to the Cr₂O₃ phase. Finally, the weak peak at 582.7 eV is likely due to CrO_xF_y species^[38]. The Cr(OH)₃ phase could be formed through the reaction of Cr₂O₃ with hydroxyl or water because the samples were exposed in air before the XPS testing.

The high-resolution F 1s core level spectra in Fig. 2 show two peaks at 685.7~686.9 and 687.4~688.6 eV for Zn/Cr₂O₃, Co/Al/Cr₂O₃, and Zn/Al/Cr₂O₃ catalysts after 118, 120, 112 h, reaction, respectively. The peak at 685.7~686.9 eV in three catalysts indicates the presence of CrF₃. The relatively weaker peak at 687.4~688.6 eV is ascribed to a new chromium fluoride phase CrO_xF_y with higher oxidation states than Cr³⁺. For Zn/Cr₂O₃ catalyst after 118 h reaction, C 1s peaks were deconvoluted into two peaks with binding energies at about 284.8 and 285.9 eV respectively, attributing to the C and -(CF₃CHCHCl)_n -. Clearly, carbon deposition over Zn/Cr₂O₃ catalyst after 118 h reaction are higher than that of the fresh Zn/Cr₂O₃ catalyst. Especially, carbon deposition derived from polymerization over Zn/Cr₂O₃ catalyst after 118 h reaction are higher than the fresh Zn/Cr₂O₃ catalyst respectively (based on the -(CF₃CHCHCl)_n -) (Fig. 2 (4) (a) and (4) (b)). The differences in Cr 2p_3/2, F 1s, and C 1s core level spectra suggest changes in the catalyst surface properties after/during the reaction, and the exterior Cr₂O₃ were strongly interacted with HF during the vapor fluorination process. Moreover, the presence of CrO_xF_y species is important to the catalytic performance (both activity and selectivity)^{[37, 39-40]}. This result was consistent with Blanchard et al^[41], observation that in a set of temperature-programmed oxidation, the fluorination activity of Cr₂O₃ is greatly influenced by the presence of Cr in higher oxidation state.

During the fluorination reaction, the Brønsted and Lewis acid site distribution of the catalyst Zn/Cr₂O₃ surface directly affect the catalytic properties of F/Cl exchange reactions^[35]. Py-IR spectrum of the catalyst Zn/Cr₂O₃ are shown in Fig. 3 (1). It can be observed from Fig. 3 that the characteristic peaks of 1466, 1504 and 1556 cm^-1 on the spectrum belong to the L acid and B acid sites respectively. The absorption peak at 1466 cm^-1 is a characteristic peak of the coordination complex of pyridine and Lewis acid site, belonging to the Lewis acid site of the catalyst. The absorption peak at 1556 cm^-1 is a characteristic peak of pyridinium ions, belonging to the Brønsted acid site of the catalyst. And the absorption peak at 1504 cm^-1 is a characteristic peak of both the coordination complex of pyridine and Lewis acid site and pyridinium ions, belonging to the Lewis and Brønsted acid site of the catalyst. Comparing Fig. 3(1) (a), (b), (c) and (d), it is easy to find that both the catalyst Zn/Cr₂O₃ after 118 h reaction and the fresh catalyst possess the Brønsted and Lewis acid site, and the numbers of acid sites of the catalyst Zn/Cr₂O₃ after 118 h reactionis significantly increased than the fresh catalyst.

The fresh catalyst Zn/Cr₂O₃ and the catalyst after 118 h reaction were characterized by the NH₃-TPD technique (Fig. 3 (2)). The NH₃-TPD profiles reveal that for the catalyst after 118 h reaction, three NH₃ desorption peaks were detected with peak temperature at 197.5, 552.9, and 702.6 ℃, respectively, corresponding to the weak and strong acid sites (Fig. 3(2), (a)), and for the fresh catalyst shows two broad desorption peaks at temperature range of 190~550 ℃(Fig. 3(2), b), which suggests that the catalyst contain mild acidic sites. The NH₃ desorption peak at 197.5 ℃ of the catalyst after 118 h, reaction is higher and wilder than the corresponding NH₃ desorption peak at 241.3 ℃ of the fresh catalyst. The two ammonia desorption peaks of the catalyst after 118 h, reaction at 552.9 and 702.6 ℃ are weaker. By contrast, one ammonia desorption peaks of the fresh catalyst at 441.9 ℃ is weaker. This indicates that acidic sites over the catalyst after 118 h reaction are more strong than those over the fresh catalyst. As noted previously, the composite catalyst Zn/Cr₂O₃ after 118 h reaction has suitable acidity sites distribution and strength, and which present high stable and selectivity under optimum fluorination conditions.

3 Conclusion

In summary, the catalyst Zn/Cr₂O₃ is the most effective catalyst among tested catalysts Zn/Al/Cr₂O₃, Co/Al/Cr₂O₃, 4% Zn/2% Cr/γ-Al₂O₃, and 8% Zn/4% Cr/γ-Al₂O_3. The addition of Zn²⁺ to Cr₂O₃ significantly increases the selectivity of HCC-1230za with HF to HCFO-1233zd (E). The heat treatment conditions in the preparation of the catalyst Zn/Cr₂O₃ has a great influence to the properties of the catalyst. When calcined at 460 ℃ under a H₂ atmosphere, it achieved at 99.4% the conversion of HCC-1230za and 98.2% the selectivity of HCFO-1233zd (E). The HCFO-1233zd (E) formation can be evidently increased with HF/HCC-1230za (molar ratio of 10:1) at 200 ℃. The results of product distribution proved the beneficial effects of molar ratio of reactants and temperature on selectivity to HCFO-1233zd (E) using Zn/Cr₂O₃ catalyst. The XRD results of the catalyst Zn/Cr₂O₃ shows that most of the amorphous chromium oxide and finely dispersed microcrystalline phase of Cr₂O₃ together lead to high activity and long lifespan. The conversion of HCC-1230za and the selectivity to HCFO-1233zd (E) are related to the specific surface area of the fluorinated catalyst. Also the higher the specific surface area of the catalyst, the higher the catalytic activity and selectivity are. The XPS spectra of the catalyst Zn/Cr₂O₃ probably indicates that the exterior Cr₂O₃ after 118 h reaction were strongly interacted with HF during the fluorination process and formed CrO_xF_y species. The number and strength of Lewis and Brønsted acid sites for the catalyst Zn/Cr₂O₃ after 118 h reaction is significantly increased than the fresh catalysts.

Acknowledgments: This work was supported by Key Research Program of Science and Technology Project of Gansu pro- vince (No.18YF1GA124).We thank Min Deng and Zhao Yanxia for our work.