Catalytic Dehydrochlorination of 1,2-Dichloroethane into Vinyl Chloride over Nitrogen-Doped Activated Carbon.

The pyrolysis of 1,2-dichloroethane (EDC) is the most popular commercialized way of producing vinyl chloride monomers (VCM); however, it is plagued by high-energy consumption and the resulting coke formation. Here, a series of nitrogen-doped (N-doped) activated carbon catalysts (N-AC) were prepared conveniently for EDC dehydrochlorination. The structural and textural properties of N-doped catalysts were characterized by X-ray diffraction, transmission electron microscopy, Raman spectra, temperature-programmed desorption of VCM and EDC, and X-ray photoelectron spectroscopy. The results revealed that doping N into activated carbon supports introduced basicity sites and caused partial graphitization on the catalyst surfaces. Thus, an improved absorption capacity to EDC and VCM and an accelerated desorption rate were obtained, which greatly enhanced EDC conversion and VCM selectivity. EDC was almost completely dehydrochlorided into vinyl chloride at a temperature of 300 °C and an EDC liquid hourly space velocity of 0.313 h-1. The high catalytic activity and selectivity as well as good stability suggested that the N-AC catalyst would be a promising dehydrochlorination catalyst on an industrial scale.

Introduction

Poly(vinyl chloride) (PVC) is one of tn class="Chemical">he most widely used plastic in the world. PVC’s monomer, vinyl chloride monomers (VCM), is commercially produced by acetylene hydrochlorination on an HgCl2 catalyst[1,2] or by pyrolysis of 1,2-dichloroethane (EDC).[3] Acetylene hydrochlorination is a main process in coal-rich countries; however, due to the high toxicity of mercury, huge environmental pressure has been exerted on this route,[4−6] especially since the implementation of Minamata Convention on Mercury.[7] EDC pyrolysis is a dominant commercialized way in the west, in which oil-based ethylene is used as the raw material. Now with the shale gas development and the decline of oil price, EDC pyrolysis has been more competitive in the preparation of VCM.

EDC pyrolysis runs at a temperature of about 500 °C and gives a reasonable conversion of 50% and VCM selectivity of 98% on an industrial scale.[8] However, tn class="Chemical">here is a problem of coke formation which is believed to be caused partly by the pyrolysis of EDC itself via acetylene through its radical chain conversion.[9,10] This problem interrupts the long-term industrial operation and requires decoking treatment every 2 months by burning coke deposit. Therefore, it is vital for EDC dehydrochlorination to develop an active, selective, and stable catalyst which can work in much lower temperatures and result in less coke deposition.

To explore suitable catalysts, a wide range of materials have been investigated for their catalytic performance in dehydrochlorination. n class="Chemical">Activated carbon (AC) supported with alkaline earth metals or transition elements were extensively studied from the 1940s. It has been firstly reported that BaCl2/AC can catalyze the dehydrochlorination of 2,3-dichlorobutane to 2-chlorobutene at 215–235 °C, but it gets deactivated in few hours.[11] Subsequently, CuCl3, CaCl2, LaCl3, and other metal chlorides supported on AC have also been extensively explored as dehydrochlorination catalysts.[12−17] Other acidic and basic catalysts, such as Al2O3, silica–alumina, ZSM-5, SAPO, polyacrylonitrile-based activated carbon fibers, and so on, have been applied in the reaction,[8,18−21] too. Moreover, phosphonium chloride ionic liquid medium has also been used as both a solvent and catalyst to the reaction in a bubbling tank reactor.[22]

In the last 10 years, our group have done much work on tn class="Chemical">he green production of VCM. We have developed an industrial transformative technology for VCM synthesis (called the J-ZH VCM process in China), in which acetylene and EDC are highly efficient and atom economically coupled to VCM over a bifunctional BaCl2/N@AC catalyst.[23−26] Now, 200 000 t/year VCM industrial demonstration line based on the new catalytic process is under construction in De Zhou, Shan Dong province, China.In the J-Zh VCM process, the elimination of HCl from EDC and the addition of HCl to acetylene occur in situ. N-Doped activated carbon catalyzes acetylene hydrochlorination and EDC dehydrochlorination at the same time.

In fact, in recent years, n class="Chemical">N-doped carbon materials have been explored as novel basic catalysts for acetylene hydrochlorination,[2,27,28] although they have also been reported as efficient catalysts for EDC dehydrochlorination, such as N-doped ordered mesoporous carbon and N-doped coconut activated carbon.[29,30] Although it has been claimed that N-doped carbon materials are effective dehydrochlorination catalysts, it is still a great challenge to develop a reliable and facial fabrication method of N-doped carbon catalysts for EDC dehydrochlorination on an industrial scale.

Based on tn class="Chemical">hese considerations and inspired by the J-ZH VCM process, in this investigation, we demonstrate a very convenient and stable method to prepare N-doped activated carbon catalysts using poly(vinylpyrrolidone) K-30 (PVP) as the N source and try to explore its performance in the elimination of hydrogen chloride from EDC as a basic catalyst. Careful characterization of the prepared N-AC catalyst, as well as systematic investigation of the activity and stability on dehydrochlorination reaction were carried out. Based on the present study, the correlation of the catalytic activities with their texture and structural properties was attempted to build. The fine catalytic performance and the available prepared method suggested that N-AC would be a potential dehydrochlorination catalyst on an industrial scale.

Results and Discussion

Structural Characterization of N-AC Catalysts

To investigate the textural properties of tn class="Chemical">he obtained catalysts, N2 adsorption–desorption isothermal analysis was performed. The N2 adsorption–desorption isotherms and the corresponding pore size distributions (PSD) on the surface of the N-AC catalysts are shown in Figures S1 and S2. The pore structure parameters of the samples are shown in Table . Clearly, all samples exhibited type-I adsorption–desorption isotherms at low partial pressures (P/P0 < 0.2) and type-IV adsorption–desorption isotherms at higher partial pressures (0.2 < P/P0 < 0.9). The results indicated the coexistence of micropores and mesopores in the catalysts. The specific surface area and total pore volume of samples decreased significantly as a function of increasing amount of PVP added. When the PVP/AC mass ratio reached 16%, the specific surface area of the sample reduced to 645.5 m2 g–1, and the pore volume was only 0.31 cm3 g–1. The results indicated that nitrogen had successfully doped within the micropores and mesopores on the AC surface altering the textural properties.

Table 1

Textual Parameters and Dehydrochlorination Performances of N-AC Samples

	N (atom %)a
samples	pyridinic	graphitic	N-oxides	P + Gc	all	SBET (m2 g–1)	Vtotal (cm3 g–1)	conversion (%)c
N-AC(0)	0.00	0.00	0.02	0.00	0.02	1058.5	0.52	31.8
N-AC(2)	0.15	0.40	0.07	0.56	0.63	975.9	0.49	50.1
N-AC(4)	0.21	0.47	0.09	0.67	0.77	976.2	0.49	57.2
N-AC(8)	0.25	0.53	0.08	0.79	0.87	845.6	0.42	60.4
N-C(16)	0.35	0.76	0.12	1.12	1.25	645.5	0.31	59.9
AM-AC	0.03	0.27	0.09	0.29	0.38	979.4	0.50	39.8
CO(NH2)2-AC	0.15	0.21	0.08	0.35	0.43	997.3	0.51	42.1
pyridine-AC	0.04	0.34	0.01	0.38	0.39	879.1	0.49	43.3
NH4Cl-AC	0.16	0.28	0.11	0.45	0.56	912.7	0.52	47.6
melamine-AC	0.12	0.34	0.05	0.46	0.51	894.2	0.46	52.7

N content was analyzed by X-ray photoelectron spectroscopy (XPS).

Reactions were carried out at 260 °C, atmospheric pressure, EDC liquid hourly space velocity (LHSV) of 0.313 h–1, VCM selectivity is above 99%.

Pyridinic and graphitic N.

XPS measurement was performed to further distinguish tn class="Chemical">he chemical valence of nitrogen on the surface of the catalysts and N bonding configuration in N-AC. As shown in Figure a, the spectra of the N-doped catalysts showed a peak at ∼400 eV, which was assigned to the N 1s photoelectron excitation and absent in AC supporter, demonstrating the successful incorporation of nitrogen into the AC surface. Moreover, the N content of N-AC, shown in Table , increased obviously with the PVP/AC mass ratio. Figure b shows the valence states of N species in the N-AC catalysts. Clearly, there were three nitrogen species coexisting in the obtained N-AC samples. The distinct peak at a binding energy of 398.6 eV was assigned to pyridinic N, the other two peaks at 400.9 and 403.2 eV were assigned to graphitic N and nitrogen oxide,[31] respectively. Noticeably, although N was originated from PVP in which N is in a C–N five membered ring, no distinctive sp3 pyrrolic-N, which should be presented at about 400.2 eV, was observed in N-AC samples.[30] This was probably because pyrrolic-N could be converted into pyridinic- and graphitic-N at the elevated carbonization temperature,[25] as shown in Figure S3. Because of similar bond lengths of C–N (1.41 Å) and C–C (1.42 Å) in sp2 hybridized hexagonal ring,[32] the N dopants in N-AC should exert less impact on the planar structure of the samples.

Figure 1

(a) XPS data of the N-AC catalysts and AC and (b) high-resolution XPS N 1s spectra of N-AC catalyst samples.

The X-ray diffrn class="Chemical">action (XRD) patterns of N-AC are shown in Figure a. Clearly, two broad peaks were observed at ∼24.6 and ∼43.1°, which were corresponding to the (002) and (100) planes of a typical graphite structure. The small peak at 26.6° should be attributed to silica, which was not completely washed away by acid-treatment in the AC supporter. The high-resolution electron micrograph of the typical catalyst sample N-AC(4) is shown in Figure b. As shown in Figure b, two different carbon structures were distinctly observed. One was amorphous carbon, the other was the crystal lattice planes associated with the typical graphite structure.[35,36] The results indicated that surface graphitization over the N-doped AC samples occurred.[33,34]

Figure 2

(a) XRD patterns of AC-N(x) samples and (b) the high-resolution electron micrograph of N-AC(4).

Raman spectra shown in Figure were performed to verify the influence of tn class="Chemical">he N-dopant on the interfacial carbon structure. Clearly, there were two explicit Raman bands. The Raman peak at about 1590 cm–1 was due to the G-band and the other peak at 1335 cm–1 was due to the D-band.[36,37] Compared with the AC supporter, the high-temperature-treated samples including N-AC(0), N-AC(4), and N-AC(8) showed narrower full width at half maximums and deeper valleys between the two bands, which indicated the structural change of graphite during the high temperature treatment process.[36] However, for these sintered samples, the intensity ratio of the D band to the G band (ID/IG) slightly increased with the PVP/AC mass ratio and the content of N dopants. Because ID/IG provides an indication of the number of structural defects and a quantitative measure of the edge plane exposure, the Raman results indicated the successful N-dopant of the N-AC samples.

Figure 3

Raman spectra of the AC parent prepared by acid treatment (a) and N-AC(x) samples, x = 0 (b), 4 (c), and 8 (d).

Catalytic Performance of N-AC

To understand the performance of tn class="Chemical">he N-AC catalyst, N doping levels, reaction temperature, and space velocity were varied. Figure a,b show the influence of N doping levels on the performance of the prepared N-AC catalysts. As shown in Figure , N-AC(0) displayed low EDC conversion (∼32%) and 96% VCM selectivity. N-Doped AC catalysts dramatically enhanced the catalytic activity with much higher EDC conversions and above 99.5% VCM selectivities. Over the obtained N-doped samples, EDC conversion firstly increased sharply with the increase of N content, and then increased slowly when the addition amount of PVP was more than 4%. Over the N-AC(8) catalyst sample, which had a total nitrogen content of 1.01%, EDC conversion reached 60%, which was even higher than that of EDC thermal pyrolysis reaction at about 500 °C. Further increasing the nitrogen content to 1.44% in sample N-AC(16), no obvious improvement in EDC conversion and a negligible effect on selectivity were observed. The above results suggested that the number of effective active centers that the reactants can reach out in the prepared N-AC samples did not increase when the amount of PVP added exceeded more than 8%.

Figure 4

Catalytic performance of N-AC catalysts. (a, b) The influence of N content. (c, d) The influence of EDC LHSV. (e, f) The influence of reaction temperature. Reactions were carried out at 260 °C, atmospheric pressure, EDC LHSV of 0.313 h–1 unless otherwise stated.

Catalytic performance of N-AC catalysts. (a, b) Tn class="Chemical">he influence of N content. (c, d) The influence of EDC LHSV. (e, f) The influence of reaction temperature. Reactions were carried out at 260 °C, atmospheric pressure, EDC LHSV of 0.313 h–1 unless otherwise stated.

In a further set of experiments, tn class="Chemical">he effect of space velocity on the reaction over the typical catalyst sample N-AC(4) was investigated and is shown in Figure c,d. As expected, EDC conversion decreased along with the rise of space velocity, whereas VCM selectivities declined slightly with the decrease of EDC LHSV. When EDC LHSV was 0.125 h–1, the conversion was about 90% and VCM selectivity was above 98.8%, and when EDC LHSV was 0.2 h–1, the conversion was about 80% and VCM selectivity was above 99.1%. When EDC LHSV increased to 0.5 h–1, the conversion was about 40% and VCM selectivity was above 99.5%. The results meant that too much long contact time would cause side reactions.

Figure e,f demonstrates the effect of the reaction temperature on the reaction over N-AC(4). Clearly, EDC conversion increased with the rise of the temperature and VCM selectivity decreased slightly when the reaction temperature was raised. When the reaction temperature was as high as 300 °C, the conversion of EDC was near 100% and the selectivity to VCM decreased slightly but was still above 98.5%. This behavior indicated the excellent catalytic performance of N-AC on the catalytic cracking of EDC into VCM.

Active Sites of N-AC

As described, the novel n class="Chemical">activity of the catalyst N-AC mainly originated from N-doping on the surface of AC. However, XPS analysis revealed that three types of N species coexisted in the obtained N-AC samples. Although great efforts have been made to identify the active sites of N-doped carbon catalysts, there are heated arguments on the role of different N specials in active sites. To confirm the active site in N-AC samples, except for N-AC, another series of catalyst samples denoted as acrylamide (AM)-AC, CO(NH2)2-AC, pyridine-AC, NH4Cl-AC, and melamine-AC were synthesized and used as dehydrochlorination catalysts, in which N was originated from acrylamide (AM), CO(NH2)2, pyridine, NH4Cl, and melamine, respectively. Because melamine is insoluble in water, methanol was used as solution instead of water. The textual parameters, N content of samples, and their performance in dehydrochlorination of EDC are summarized in Table , too. From Table we can see that all of these samples showed similar textual parameters. The surface area varied slightly from 900 to 1000 m2 g–1 and the total pore volume was 0.46–0.52 cm3 g–1, so the effect of textural properties on the catalytic performance could be ignored. XPS measurements demonstrated the successful incorporation of nitrogen into the AC. Similar to N-AC, there were three N species including pyridinic N, graphitic N, and nitrogen oxide N. The total N content varied from 0.3 to 0.8 atom % in these samples, and the distributions of N species in the samples were different. Meanwhile, the performance of these samples on EDC dehydrochlorination is listed in Table , too. Clearly, the inconsistent result of the EDC conversion and the total N content was observed. As noted from Table , the total N content of NH4Cl-AC samples was the highest among them, but EDC conversion over it was lower than that over melamine-AC.

So we correlated the catalytic properties with tn class="Chemical">heir nitrogen species that originated from all nitrogenous compounds including PVP, AM, CO(NH2)2, pyridine, NH4Cl, and melamine. Because the Brunauer–Emmett–Teller (BET) surface of AC-N(16) was much smaller than the others, it was not considered in the fitting process. Thus, the difference of the catalytic performances of all samples including N-AC(0) was mainly caused by different amounts of N species in the samples. The result is shown in Figure . As shown in Figure , the EDC conversion linearly increased with the increasing amount of both pyridinic N and graphitic N (y = 37.63x + 30.91), irrespective of the varying N resources, inferring that pyridinic N and graphitic N were closely correlated with the catalytic dehydrochlorination of EDC and were effective N atoms.

Figure 5

Fitting between both pyridinic N and graphitic N species and reactivity towards dehydrochlorination.

It is generally known that pyridinic and graphitic-N refer to n class="Gene">sp2 hybridized nitrogen atoms bonding with C atoms in the hexagonal ring.[38,39] Pyridinic N contributes three electrons to fill two σ-bonds and one aromatic π-bond and leave a lone electron pair in the plane of the carbon matrix, whereas graphitic-N contributes four valence electrons forming three σ-bonds and one π-bond with the neighboring carbon (or hydrogen) atoms, and the fifth electron is shared by the graphitic-N dopant itself and the π* state of the conduction band. In the case of the prepared N-AC catalysts, pyridinic-N and graphitic-N species increased the electron–donor property of the activated carbon and thus introduced basic centers into the carbon material, so the derived N-AC could be used as a solid-base catalyst. Although it was reported that pyridinic and pyrrolic N species were critical for the reaction but quaternary N species caused catalyst deactivation,[30] the above results verified that both pyridinic N and graphitic N introduced basic sites and made for catalytic dehydrochlorination of EDC.

The adsorptive properties of n class="Chemical">EDC and VCM on N-AC(4) and N-AC(0) were further studied by temperature-programmed desorption (TPD) analysis. The results are shown in Figure . Clearly, the peak areas of N-AC(4) in both the EDC-TPD profile and VCM-TPD profile were much larger than those of N-AC(0), suggesting the enhanced adsorption capacity in N-AC(4) which should be associated with the introduced basic centers verified by XPS results. However, it was worth noting that the peak positions of N-AC(4) in the VCM-TPD profile and EDC-TPD profile were 106.6 and 171.7 °C, respectively, whereas the corresponding peaks of N-AC(0) was 110.2 °C in the VCM-TPD profile and was 188.6 °C in the EDC-TPD profile. The much lower desorption temperatures of N-AC suggested that both reactant EDC and product VCM could be easier and faster to dissociate from the N-doped AC surface. The distinguished adsorption properties of N-AC should not only make for higher EDC conversion, but also avoid deep side reactions caused by excessive contact time and favor better VCM selectivity.

Figure 6

TPD profiles of the catalysts. (a) 1,2-EDC-TPD and (b) VCM-TPD.

Stability and Deactivation of N-AC

To explore the stability of n class="Gene">N-AC samples, a series of experiments were carried out on the typical catalyst sample N-AC(4), and the results are shown in Figure . Figure a shows the effect of EDC LHSV on the stability of N-AC(4) at the reaction temperature of 260 °C. As demonstrated in Figure a, the conversion of EDC over N-AC(4) was 95% at an EDC LHSV of 0.125 h–1. After 10 h, EDC LHSV was raised to 0.313 h–1, whereas the other conditions were kept constant; here, the corresponding conversion of EDC lowered to 67%. After maintaining this condition for 12 h, the EDC LHSV was readjusted to 0.125 h–1; here, the conversion of EDC recovered to about 94%, which was very close to the initial lever. Meanwhile, VCM selectivity was consistent and stable in the whole process.

Figure 7

Stability of catalyst N-AC. (a) The influence of EDC LHSV at 260 °C. (b) The influence of reaction temperature at EDC LHSV of 0.313 h–1. (c) Long-time stability test over N-AC(4) at 260 °C and EDC LHSV of 0.313 h–1.

Figure b shows the effect of tn class="Chemical">he reaction temperature on the stability of N-AC(4) at an EDC LHSV of 0.313 h–1. As shown in Figure b, EDC conversion was 37% at a temperature of 250 °C. After 10 h, the reaction temperature was raised to 290 °C and the corresponding EDC conversion was about 90%. When the reaction temperature was readjusted to 250 °C 10 h later, EDC conversion returned back to about 38%. Meanwhile, VCM selectivity in all cases were relatively static and above 99%. The results indicated that the intermittent changes of a space velocity and reaction temperature had little influence on the catalyst activities, which was very critical for the industrial catalyst.

To explore the long-term stability and den class="Chemical">activation of the prepared N-AC catalyst, the amplified single tube test (φ = 30 mm, EDC LHSV 0.313 h–1) was carried out over 120 g N-AC(4), and the result is exhibited in Figure c. As can be noted from Figure c, the EDC conversion was half of the initial one after 560 h, whereas the VCM selectivity was above 99% at all times. The results suggested the good stability of the N-AC catalyst. To explore the cause of deactivation, the used N-AC(4) was carefully characterized by XPS measurement, N2 gas adsorption and thermogravimetric (TG) analysis.

As shown in Table , N content in the used n class="Gene">N-AC(4) decreased greatly. The total N content downed from 0.89% in the fresh sample to 0.21%. The further deconvolution of the XPS N 1s profile proved that the used N-AC(4) retained three N species including pyridinic-N, graphitic-N, and N-oxides, and no other N species was produced in the reaction. Moreover, the proportions of pyridinic-N, graphitic-N, and N-oxides were, respectively, 28.57, 57.14, and 14.29%, which were very close to the N species distribution in the fresh AC-N(4) (26.97, 60.67, and 12.36%). No obvious change of the nitrogen species distribution was observed before and after the catalytic reaction. The results suggested that N species in the sample were very stable, and no obvious transformation of nitrogen species to each other occurred. The reduction of nitrogen content, especially pyridinic-N and graphitic-N, rather than the transformation of nitrogen species,[30] contributed to partial deactivation of the catalyst.

Table 2

Textual Parameters and Elemental Composition of the Fresh and Used N-AC(4) Catalyst

	N (atom %)a
samples	all	pyridinic	graphitic	N-oxides	Cl	SBET (m2 g–1)	Vtotal (cm3 g–1)
fresh	0.77	0.21	0.47	0.09	0.00	976.2	0.49
used	0.18	0.05	0.10	0.03	0.35	260.8	0.02

N and Cl content were analyzed by XPS.

In contrast to the decrease in n class="Chemical">nitrogen content, chlorine content in the spent sample increased greatly. As can be noted in Table , a large amount of chlorine occurred in the used catalyst, whereas only a trace amount was checked in the fresh sample, which should be caused by hydrochloric acid pretreatment. The further convolution of Cl 2p3/2 and 2p1/2 shown in Figure c suggested that some chlorinated organic compounds such as chlorobenzenes and poly(vinyl chloride) covered the surface of the used catalyst, which was further approved by the high-resolution electron micrographs shown in Figure .

Figure 8

(a) XPS N 1s spectrum of fresh N-AC(4) and (b) used N-AC(4), (c) XPS Cl 2p3/2 and p1/2 of the fresh and used N-AC(4), and (d) TGA results of fresh and used N-AC(4).

Figure 9

High-resolution electron micrographs of coke deposited on the used N-AC(4).

The textural properties of tn class="Chemical">he used N-AC(4) were further estimated and are also shown in Table . As listed in Table , the surface area of the used N-AC(4) decreased from 976.2 to 260.8 m2 g–1 and the total pore volume downed from 0.49 to 0.02 cm3 g–1, suggesting that the catalyst pores were partly blocked after the catalytic reaction. Further TG analysis under a N2 atmosphere shown in Figure d confirmed much higher mass loss in the temperature range of 250–650 °C compared with the fresh one. Without doubt, the extra part of weightlessness should be attributed to the chemicals such as coke deposited on the catalyst surface which caused the decrease of the surface area and pore volume after the reaction. Meanwhile, the deposited chemicals covered the surface of the catalyst and caused the decreased N specials. Thus, it could be concluded that coke deposition, that caused the pore blocking and N content decrease within the micropores and mesopores on the AC surface, was the main reason for the partial inactivation of the N-AC catalyst.

Conclusions

In conclusion, we have demonstrated here a series of n class="Gene">N-AC catalysts prepared by a simple and reliable method and their excellent performance in EDC dehydrochlorination. The results revealed that doping N into AC supports introduced surface-basicity sites and caused partial graphitization on the catalyst surfaces. Further experimental results indicated that both pyridinic N and graphitic N introduced basic sites for catalytic dehydrochlorination of EDC. The high catalytic activity and selectivity as well as good stability suggested N-AC to be a promising dehydrochlorination catalyst on an industrial scale.

Experimental Section

Raw Materials

In our research, columnar commercial coal-based n class="Chemical">activated carbon (AC) was supplied by Ningxia Bethel Activated Carbon Co., Ltd. (Ningxia, China), poly(vinylpyrrolidone) K-30 (PVP), urea (CO(NH2)2), acrylamide (AM), pyridine, NH4Cl, melamine, methanol, hydrochloric acid (37.5 wt % aqueous solution), and EDC were purchased from Titan Technology Co., Ltd. (Shanghai, China). All of these commercial guaranteed reagents (GR grade) were used without further purification.

N-AC Catalyst Preparation

In our research, N-AC catalyst samples were prepared by an incipient wet impregnation technique using columnar commercial n class="Chemical">AC as the carrier and PVP as the N source. First, columnar commercial AC was washed with dilute aqueous hydrogen chloride (2 mol L–1) at 50 °C overnight to remove metal ions, which would affect the catalytic dehydrochlorination reaction, and then was filtered by pure water five times and dried 120 °C overnight. Typically, 100 g of acid-washed AC was added to 150 mL of PVP aqueous solution (the mass ratios of PVP to AC is 2, 4, 8, and 16%), and sonicated for 5 min. The mixture was incubated at 60 °C for 10 h and dried at 100 °C for 24 h. After being sintered for 6 h at 650 °C with a heating rate of 2 °C min–1 under the protection of N2, the N-AC samples were obtained and denoted as N-AC(2), N-AC(4), N-AC(8), and N-AC(16). N-AC(0) was obtained from the N-AC parent by sintering in N2 for 6 h without N doping. To confirm the active site in N-AC samples, another series of catalyst samples denoted as AM-AC, CO(NH2)2-AC, pyridine-AC, NH4Cl-AC, and melamine-AC were synthesized by the same method, in which N originated from acrylamide, urea, pyridine, NH4Cl, and melamine, respectively. Because melamine was insoluble in water, methanol was used as a solution instead of water when the melamine-AC sample was prepared.

Characterization of Catalyst Samples

Powder X-ray diffraction (XRD) patterns of various syntn class="Chemical">hesized samples were recorded on a D8 Advance Da Vinci with a Cu source (40 kV, 20 mA). The high-resolution transmission electron micrograph was operated at an FEI Talos F200X at 300 kV. N2 adsorption at −196 °C was carried out on a Micromeritics ASAP 2460 apparatus. XPS was carried out on a Kratos AXIS ULTRA DLD spectroscope with Al Kα X-rays as the excitation source at 1000 meV and 150 W. The binding energies were calibrated by the contaminant carbon (C 1s 284.6 eV). The pressed powder pellet method was used to minimize the analytic error, which would be caused by the limit of the electron escape depth based on XPS analysis. Raman spectra were obtained with a Senterra R200-L using a 514.5 nm AR line as an excitation source. Thermogravimetric (TG) analysis was carried out on TA Instruments-Waters LLC Q-600 STD, in which the temperature was ramped to 800 °C at a heating rate of 10 °C min–1. TPD was performed on a Micromeritics AutoChem II 2920 equipped with a thermal conductivity detector in the temperature range of 50–600 °C. The heating rate was 10 °C min–1, and the gas flow rate was 20 mL min–1. Temperature-programmed desorption (TPD) experiments were initiated by pretreatment of the catalyst with He for 30 min at 600 °C. Subsequently, the catalyst temperature was cooled down to 150 °C and an adsorption gas was introduced for 20 min. The catalyst temperature was then kept constant for 20 min to flush away any loosely bound surface species. Subsequently, the catalyst temperature was decreased to 50 °C. The step of desorption was then carried out in the temperature range of 50–600 °C.

Catalytic Reactions

Dehydrochlorination of EDC was carried out in a fixed-bed quartz tube ren class="Chemical">actor (diameter = 14 mm, length = 700 mm) at atmospheric pressure and 220–300 °C. Generally, 15 g of catalyst was used. The pipeline and quartz tube reactor were flushed with nitrogen to remove water and air from the system before the reaction. Pure EDC was then fed into the entrance of the reactor by a double-plunger micro-pump, and then vaporized. The reactant and products were analyzed by a gas chromatograph (GC) (Shimadzu GC-2014), and HCl was analyzed by acid base titration.