Nitrogen Modified Carbon Nano-Materials as ... - ACS Publications

Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Nitrogen Modified Carbon Nano-Materials as Stable Catalysts for Phosgene Synthesis Navneet Kumar Gupta, Bo Peng, Gary L. Haller, Erika Elisabeta Ember, and Johannes A. Lercher ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01424 • Publication Date (Web): 11 Jul 2016 Downloaded from http://pubs.acs.org on July 15, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Nitrogen Modified Carbon Nano-Materials as Stable Catalysts for Phosgene Synthesis Navneet K. Gupta, Bo Peng, Gary L. Haller, Erika E. Ember* and Johannes A. Lercher* Technische Universität München, Department Chemie, Lichtenbergstr. 4, D-85747 Garching, Germany.

ABSTRACT. The carbon catalyzed reaction of Cl2 and CO constitutes the most important industrial route to phosgene. While defects in carbon lead to surface chemical reactions, direct polarization of C-heteroatom bonds induces a more successful Cl2 catalytic activation, the rate determining step in the overall catalytic cycle. The interplay between the electron donating and withdrawing ability of the incorporated nitrogen substituents on the formation and stabilization of active sites was examined by X-ray photoelectron and Raman spectroscopy. Mechanistic studies indicate that the polarized Cl2 induced by the direct interaction of Cl2 with a strongly electron deficient carbon site in close proximity to a nitrogen substituent is essential for phosgene production. Nitrogen substitution into ordered carbon materials led to very active and stable carbon catalysts for COCl2 synthesis.

KEYWORDS. Cl2 activation, transient COCl2 synthesis, N-containing carbon catalyst, electrophilic and nucleophilic carbon sites, in situ Raman spectroscopy, reaction mechanisms

ACS Paragon Plus Environment

1

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 39

1. INTRODUCTION Carbon based materials have received increasing attention as metal-free catalysts,1-3 because different hybridizations of carbon lead to strain surface structures4 and enables diverse modes of adsorption.5,6,7 It has been shown that insertion of heteroatoms, such as N, O, P, and B, allows controlled variation of the electronic properties of carbon sites.8-11 In particular, nitrogen has proven to be successful for carbon modifications, leading to a broad variety of nitrogen-modified materials, including N-doped carbon nanotubes,12,13,14 colloidal graphene quantum dots15 and carbon sub-micrometer spheres.16 Several studies have suggested that N influences the spin density and charge distribution on neighboring carbon atoms17,18 enhancing their catalytic activity.19,20,21 Generally, nitrogen containing heterocyclic rings consist of three main nitrogen types, i.e., pyridinic, pyrrolic and quaternary nitrogen. The electronic structure of the adjacent carbons depends on the electronic environment of N and the nature of the C-N bonding configurations. Pyridinic N possesses a strong electron affinity and creates a high net positive charge density on the adjacent carbon atoms18,22, while pyrrolic nitrogen induces a high electron density (negative charge) on the vicinal carbons.23 Quaternary-N substitution in aromatic molecules stabilizes carbenes by mesomeric and inductive synergetic effects.24 The specific impact of these N modifications on catalysis is still discussed controversially. As one of the potential applications of such chemically modified carbon, we have explored in this contribution the role of nitrogen presence for carbon catalyzed synthesis of COCl2 from Cl2 and CO. In a recent report, the presence of bent carbon hemispheres and cages has been shown to be important, as the strain on the sp2 hybridized C enhances the ability of triplet states to adsorb and activate Cl2.25 The interaction of Cl2 with such carbons resulted in the formation of an activated Cl2 with radical character. The subsequent attack of physisorbed


2

Page 3 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

CO resulted in the transient formation of [COCl·], which further reacts to COCl2 via Cl2 insertion.25 Conventional activated carbon catalysts degrade during COCl2 synthesis by attack of Cl2 and Cl· on carbon defects, which lead to the corrosive formation of CCl4.26 Thus, defectuous planar carbon surfaces are, therefore, inactive for COCl2 formation, but Cl2 corrosively reacts at their terminations boundaries. The present paper explores the role of N in that catalytic chemistry, changing the nature of C via inductive as well as mesomeric electronic interactions. The impact of N-modifications on the generation of active sites in carbon materials for stable and selective COCl2 synthesis is described.

2. EXPERIMENTAL 2.1 Nitrogen containing carbon materials The N modified mesoporous carbon materials were prepared by published procedures.27 The proposed polymeric structures are presented in Figure 1.


3

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 39

Figure 1. Proposed structures of N-containing mesoporous ordered carbon materials synthesized from 1,4-dicyanobenzene (a, Polymer A), tetrachloroterephthalonitrile (b, Polymer B), 9,10-anthracenedicarbonitrile (c, Polymer C) and 4,4′-biphenyldicarbonitrile (d, Polymer D). For the preparation of 1,3,5-triazine based polymer A shown in Figure 1, a mixture of 1,4dicyanobenzene (7.8 mmol, 98 % from Sigma Aldrich) and anhydrous ZnCl2 (5 mole equivalents, 97 % from Sigma Aldrich) were sealed in a glass ampule under inert conditions. The ampule was heated to 673 K with 1 K/min and maintained at this temperature for 40 h. Then, the ampule was cooled to ambient temperature and carefully opened (high pressure in the ampule resulted from the thermal generation of gaseous degradation products). The solid remainder was ground and washed thoroughly with small aliquots of bi-distilled H2O and dilute HCl. The resulting black powder was refluxed at 373 K in 1 M HCl for 48 h to quantitatively remove the remaining ZnCl2. Finally, the polymer was further purified by


4

Page 5 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

washing with 1 L of bi-distilled H2O followed by THF (99.9 %, Sigma Aldrich, 15 ml). The resulting solid was dried in vacuum at 350 K overnight. To understand the influence of structural variants of N-substituted carbons, also 9,10anthracenedicarbonitrile (97 % from Sigma Aldrich), and 4,4′-biphenyldicarbonitrile (97 % from Sigma Aldrich) were reacted as building units. These precursors yield 1,3,5-triazine rings interconnected with different aromatic spacers (biphenyl and anthracene, respectively) resulting in texturally different carbons. Using tetrachloroterephthalonitrile (95 % from Sigma Aldrich) as reagent led to a mesoporous material (Figure 1, structure B) where all sp2 carbon atoms are blocked by covalently bonded Cl. In a comparative study of Cl2 activation, pyridinic (type A) and quaternary nitrogen containing molecules (type B) were explored (Figures 2 and 3). The molecules with a purity higher than 97 % were purchased from Sigma Aldrich and were used without further purification.

Figure 2. Overview of studied pyridinic nitrogen containing molecules (Type A): pyrazine (a), 1,3,5-triazine (b), melamine (c) cyanuric chloride (d), 2,6-dichloropyridine (e), 2,4,6trichoropyridine (f) and 2,6-dichloropyridine N-oxide (g).


5

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 39

Figure 3. Overview of studied quaternary nitrogen containing molecules (Type B): 1,3bis(2,4,6-trimethylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene (a, BTDIY), 1,3-bis(2,4,6trimethylphenyl) imidazolium chloride (b, BTIC), 1-Ethyl-2,3-dimethylimidazoliumtrifluoromethanesulfonate (c, EDITM).

2.2 Catalytic Cl2/CO activation and COCl2 synthesis Cl2 activation on all synthesized N-containing carbons along with selected well-defined nitrogen containing substances was studied between 300 and 473 K. The materials were compared to high surface area graphite (HSAG, specific surface area of 300 m2/g, received from Timcal®). For each experiment 10 − 100 mg of solid catalyst (particle size ≤ 80 µm) was placed in a quartz reactor (0.4 cm ID) and thermally activated at 423 K and 1 bar for 1 h prior to multiple Cl2 pulses (1.8 µmol Cl2 / pulse, pulse length 230 s) in He (10 cm3 STP per min) were applied. All gases were analyzed by on-line mass spectrometry (OMNIstar/TM mass-spectrometer). For the identification of the surface bound species, temperature programmed desorption (TPD) was performed from 300 to 993 K. CO interaction on fresh and Cl2 exposed catalysts was also studied under conditions similar to Cl2 activation. To test the catalytic performance for COCl2 synthesis, the catalyst was exposed to repetitive Cl2 pulses in 5 vol. % CO in He (10 cm3 STP per min). 2.3 Experimental set-up and safety issues A schematic representation of the experimental setup constructed for the Cl2 activation and CO interaction study is shown in the supporting material, Figure S1. The reactor system allowed to operate with minimal amounts of hazardous gases. A six-port valve connected to


6

Page 7 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

an on-line mass spectrometer (MS) was used for the automated dosage of Cl2. All gaseous effluents were neutralized in two interconnected washing bottles containing concentrated KOH and Cu2+ solution. An active carbon containing filter retained the remaining toxic gaseous intermediates. Cl2 and CO detectors were placed in the fume hood. To avoid exposure to CO, two additional filters for CO entrapment were installed at the outlets of the six-port valve and of the reactor. The maximum amount of COCl2, which could be formed in the pulse reactor (0.025 ppm) was lower than the exposure limit (0.1 ppm averaged over a work shift of up to 10 hours a day, 40 hours per week with a ceiling level of 0.2 ppm averaged over a 15 min period).28 2.4 Catalyst characterizations 2.4.1 BET surface area Nitrogen physisorption was measured on a PMI automated Sorptomatic 1990 instrument operated at liquid N2 temperature (77 K). Prior to the experiments, 200 mg of the Ncontaining C samples were outgassed and activated in vacuum (10-4 mbar) at 473 K for 2 h. After activation, the samples were weighed and cooled to 77 K. Specific surface areas were estimated by applying the BET-theory (p/p0 from 0.05 to 0.25). The pore size distributions (PSD) were obtained by applying the Barret–Joyner–Halenda (BJH) model29 to the adsorption branch of the isotherm. Pore volumes were evaluated from the αs comparative plot with nonporous hydroxylated silica as the reference adsorbent.30 The t-plot method was used to differentiate micro- from mesoporosity.31 2.4.2 Elemental analysis (EA)


7

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 39

Elemental compositions of the materials were determined by CHN analysis at TUMünchen. The quantity of oxygen was estimated at Mikroanalytisches Labor Pascher, Germany. 2.4.3 Scanning electron microscope (SEM) Topographical imaging of the materials was performed using a JEOL JSM-5900LV and Hitachi S3500N SEM. The SEM was operated at acceleration voltages between 10 and 25 kV, using a working distance of 9 mm for imaging. 2.4.4 Raman spectroscopy The Raman measurements were performed with a Renishaw Raman spectrometer Series 1000. After calibration using an Si(111) single crystal the Raman spectra were recorded using the green line of an Ar-ion laser (514.53 nm, 2.41 eV) at elevated temperature under inert conditions. For ex situ measurement the thermally activated (423 K, 1h) samples were prepared over glass slits. The final spectra were 1-10 accumulations in a range of 100−4000 cm-1. Similar conditions were applied for in situ Raman measurements, performed by placing materials between quartz wool in a quartz reactor of diameter 1.62 mm ID. Spectra were recorded after thermal induced desorption of small molecules from the carbon surfaces (N2 flow at 423 K and 1 h). Subsequently, Cl2 was admitted to the carbon at elevated temperature, until the Raman spectrum of carbon interacting with Cl2 did not change. The Cl2 reversibility on the surface was checked by switching back to inert gas (N2) flow. 2.4.5 ESR spectroscopy ESR spectra of materials were recorded in the perpendicular mode on an X-band Joel Jes Fa 200 spectrometer equipped with a cylindrical mode cavity and a liquid He cryostat. The ESR measurements were performed, by taking spectra of samples inside a quartz tube in a


8

Page 9 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

temperature range of 298−473 K using 9.45 GHz, 1 mW microwave power. The analysis and simulation of the data was done with the Jes-Fa Series software package Version 2.2.0.

2.4.6 X-ray photoelectron spectroscopy (XPS) For the quantitative analysis of the surface chemical composition and bonding, X-ray photoelectron spectroscopic measurements were performed on ULVAC-PHI, Type Versa Probe 5000 spectrometer using monochromatic Al Kα radiation. All the samples were ground (particle size < 80 µm) and mounted on a double sided adhesive tape before introduction in the measuring chamber. The detection depth is 5 to 10 nm and the identification limit is around 0.1 atom %. The overall XPS spectra was recorded with a 187.8 eV pass energy analyzer and an increment of 1eV. High-resolution individual spectra were obtained by using a 23.5 eV pass energy and an increment of 0.1 eV. The data analysis has been performed by using MultiPak software version 9.4.1.2. All the recorded spectra were calibrated to a 284.8 eV C1s binding energy. The binding configuration of the studied elements were identified based on the chemical shifts in the photoemission spectra and by a direct comparison with the reported literature values.32 The XPS peaks were fitted to Voigt functions having 60 % Gaussian and 40 % Lorentzian character, after performing a Shirley background subtraction. Deconvolution of the XPS spectra was done by applying fixed binding energy values for the different N- and C- types and allowing the full width at half maximum (lower limit-1.15 eV and upper limit-1.65 eV) to vary in order to obtain the best possible fit. 3. RESULTS AND DISCUSSION 3.1 Physicochemical properties of materials


9

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 39

A series of porous nitrogen-containing carbon materials (Figure 1) was synthesized via a Zn2+ acid catalyzed one-pot reaction. In a typical polymerization well-ordered layered materials were formed containing N- and O-based substituents on the aromatic rings at terminal positions (Table S1 and Figure S2). The textural properties of synthesized materials varied markedly with the precursors (Table S3). The porous 1,3,5-triazine based polymers from 1,4-dicyanobenzene and 9,10-anthracenedicarbonitrile had similar specific surface areas (775 - 746 m2/g) and a broad pore size distribution. The cross-linked networks in the polymer A (Figure 1) synthesized from 1,4-dicyanobenzene had a significantly higher micropore volume (Vmicropore = 0.374 cm3/g) compared to the material synthesized by using 9,10anthracenedicarbonitrile (Vmicropore = 0.063 cm3/g). However, the use of carbon with the bulkier anthracene group as spacer resulted in the increase of mesoporosity from Vmesopore = 0.096 cm3/g (for benzene) compared to 0.429 cm3/g (for 9,10-anthracenedicarbonitrile). We hypothesize that the difference is caused by variations in planarity of anthracene molecules located between the triazine rings (for further information, see SI Figure S3). In contrast, the use of tetrachloroterephthalonitrile resulted in notable reduction of the specific surface area of the triazine based polymer A from 775 to 285 m2/g and a considerable decrease in the material porosity (Vmesopore = 0.096 to 0.026 cm3/g, Vmicropore = 0.374 to 0.012 cm3/g, respectively). The decrease in surface area, mesoporosity and microporosity were attributed to the presence of covalent C-Cl bonds on the benzene linker located between triazine groups. HR-SEM images of the materials obtained from tetrachloroterephthalonitrile showed a three dimensional layered structure of the materials similar to the case of the material resulting from 1,4-dicyanobenzene (Figure S3 (a) and (b)). The carbon material with the highest specific surface area and mesoporosity was obtained by using 4,4′-biphenyldicarbonitrile as starting precursor (1558 m2/g). It is hypothesized that the larger distance between the -C≡N


10

Page 11 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

groups of the 4,4’-bisphenyldicarbonitrile compared to 1,4-dicyanobenzene materials allows to stabilize the high specific surface area. The Raman spectra of the carbon in the synthesized materials showed two distinct bands at 1350 cm-1 (D-band) and 1580 cm-1 (G-band), corresponding to the defect amorphous and the well-ordered crystalline phase (Figure S4).33,34 For analysis of the well-ordered graphitic carbon and the surface structural defects, the intensity ratios of D- to G-bands (ID/IG) were estimated by using the method published by Sadezky et al. and Haghseresht et al. and the deconvoluted spectra of Figure S5.35,36 The ID/IG ratio (intensity ratio of D1 and G bands, Table S4) of synthesized materials ranged from 0.75 (for the 1,3,5-triazine based polymer D synthesized from 4,4′-biphenyldicarbonitrile) to 1.55 (for the polymer B obtained from tetrachloroterephthalonitrile, shown in Figure S5 and Table S5). Comparing the structural properties of the N-containing carbons to HSAG (ID/IG = 0.28, Table S5) shows a higher concentration of structural defects in the nitrogen containing carbon materials. In contrast to triazine-based polymers, the N-containing small molecules had narrow vibrational bands (Figure S8), but lacked the collective vibrations of extended solids (phonon modes). The chemical compositions of the synthesized carbons varied by changing the precursors (additional information in Table S1 and S2). Substantial variations in the C/N ratios from 3 for the polymeric material B (ex tetrachloroterephthalonitrile) to 21.3 for the polymeric material C (ex 9,10-anthracenedicarbonitrile) were observed. A more accentuated increase in the chlorine content (6.81 at. %) was observed for the material synthesized from tetrachloroterephthalonitrile (Supporting Information, Table S1 and S2). High temperature acid hydrolysis of nitrile groups resulted in oxygen incorporation in all carbon materials at the level of 3.7 to 8.1 at. % oxygen containing functional groups.


11

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 39

The deconvoluted XPS spectra of N1s binding state (Figure S9) indicate that N exists in six different chemical environments in the studied N-containing carbons, i.e., pyridinic-N (BE = 398.62 ± 0.11 eV), amine (BE = 399.83 ± 0.10 eV), pyrrolic-N (BE = 400.62 ± 0.11 eV), quaternary nitrogen (BE = 401.82 ± 0.11 eV), N-oxides of pyridinic-N (BE = 403.32 ± 0.11 eV) and NOx (BE = 404.72 ± 0.11 eV).37,38 The relative content of each N bonding configuration is summarized in Table 1 for the four triazine based polymers shown in Figure 1. Table 1. XPS analysis of N1s binding state for synthesized different N-containing carbons. N 1s binding states (%) Precursor Pyridine

Pyrrole/ Amine Quart. Pyr.-NO -NOx amide /nitrile

Atom % of N

1,4-Dicyanobenzene

55.3

31.2

9.6

0

4.0

0

10.6

Tetrachloroterephthalonitrile

62.6

27.2

9.6

0.3

0

0.2

14.7

9,10-Anthracenedicarbonitrile

53.6

25.3

11.4

3.2

5.6

1.0

3.6

4,4`-Biphenyldicarbonitrile

40.5

38.0

15.4

3.4

2.8

0

5.5

The deconvolution and quantification of XPS spectra showed that the relative distribution of the N bonding configurations is similar in the four materials. Thus, in all cases the most abundant N-function is pyridinic-N, with 40.5-62.6 % of all N, followed by pyrrolic-N and/or amides (25.3-38 %) and amine/nitrile functionality (9.6-15.4 %). Aromatic sp2 hybridized N, such as in pyridine, leads to substantial electron deficiency in the aromatic ring.39 In contrast, the N-atom in an amine function enhances the electron density of the ring. Materials from 1,4-dicyanobenzene and tetrachloroterephthalonitrile contain about 9.6 ± 0.4 % amine/nitrile functionalities. This was slightly higher (11.4 - 15.4 %) in materials synthesized from 9,10anthracenedicarbonitrile and 4,4`-biphenyldicarbonitrile. It should be noted, that based on the


12

Page 13 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

XPS analysis, a clear distinction between the aromatic amine and nitrile functionalities was not possible.44 Pyrrolic-N refers to N atoms that are bonded to two carbon atoms and contribute to the aromatic system with two π-electrons. For 9,10-anthracenedicarbonitrile and tetrachloroterephthalonitrile 27 and 25 % pyrrolic-N were found, respectively, i.e., 1.2 -1.5 times higher than concentrations found in materials synthesized from 1,4-dicyanobenzene and 4,4`-biphenyldicarbonitrile. With the exception of the material synthesized from 1,4dicyanobenzene, all other samples contained 0.3-3.4 % of quaternary N (Table 1), which replaces a C-atom in the polymer layers. N-oxides (0.2 to 1.0 %) were also found in 9,10anthracenedicarbonitrile and tetrachloroterephthalonitrile. Depending on the electronic impact of the N modifications on the neighboring carbon atoms, electron donation and withdrawal have been observed (Table 2). Table 2. Classification of N-functional groups based on their electronic effect (Rresonance/mesomeric, I-inductive) towards aromatic C.

Electron withdrawing effect (-I/-R )

Electron donating effect (+I/+R)


13

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 39

The electronegative pyridinic-N induces a negative mesomeric effect (-R effect) and stabilizes the electrophilic sp2-carbon site.40 In contrast, the amine and pyrrolic-N groups promote a more accentuated positive mesomeric effect (+R) rather than an -I effect and, therefore, lead to higher electron density (nucleophilicity) on the adjacent carbon of an aromatic ring. In the case of pyridinic-N oxides the +I effect dominates and contributes to the stabilization of the negative charge on highly electronegative oxygen atom. The quaternary-N modification has a –I effect and significantly contribute to the stabilization of electrophilic C.41 Additional substitution promotes the formation of stable carbenes.24 The impact of these N atoms on the overall state of the carbon was explored C1s X-ray photoemission (Figure S10). The deconvoluted XPS spectra are shown in supporting information Figure S10 and binding energies are summarized in Table 3. Table 3. XPS analysis of C1s binding state for synthesized different N-containing carbon materials. C1s binding states (%) Precursor

1,4-Dicyanobenzene

-C=C-C-H 284.5 eV 68.1

Ph-Cl -C-O-C=O -O-C=O -C=N-O-C-O285.5 eV 286.15 eV 289.55 eV 288.65 eV 9.1 13.7 3.8 5.2

Tetrachloroterephthalonitrile

59.2

20.8

14.7

3.5

1.9

9,10-Anthracenedicarbonitrile

66.6

9.8

14.5

1.7

7.4

4,4`-Biphenyldicarbonitrile

72.8

7.4

12.0

0.4

7.4

The main peak at 284.5±0.31 eV, attributed to the graphitic C.40 The peak at 285.5 ± 0.2 eV is assigned to the combination of aryl carbon bonded to –Cl and / or double bonded to N,43 whereas the peaks at 286.15 ± 0.25, 287.55 ± 0.31 and 288.65 ± 0.26 eV are related to carbon


14

Page 15 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

atoms attached to different O-moieties in the form of ether/alcohol, ketone, and carboxylic acids, respectively.44,45 Quantitative analysis of the XPS data showed the presence of 59-73 % of graphitic C in all N-containing carbon materials. For carbon derived from tetrachloroterephthalonitrile, the sp2 graphitic C contribution decreased 10 % and the contribution from the C–Cl groups increased by the same amount. The -Cl incorporation (1-2 wt. %) in other carbons is concluded to be caused by the ZnCl2 as catalyst. The aryl-OH (12-14.7 %), C=O (≤ 3.8 %) and RCOOH (carboxylic acid) (≤ 7.4 %) groups in all materials were due to acid catalyzed hydrolysis reactions at high temperature and pressure in the synthesis ampule.46,47 3.2 Cl2 activation 3.2.1 On-line MS analysis of the catalytic reaction course The Cl2 activation was studied using pulsed admission of the reactant in a quartz reactor (supplemental material, Figure S1). After thermal activation of the selected carbon materials, during the first Cl2 pulses, HCl (m/z: 36) was released from all samples (Figure 4). In that time no molecular Cl2 (m/z: 70) was detected in the gas phase, indicating that irreversible surface processes occurred, which consumed Cl2 (Equation 1 and 2).


15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Gas phase products (µmol)

ACS Catalysis

Page 16 of 39

Cl2 (m/z: 70)

Cl2 pulse

1,2

HCl (m/z: 36)

1,0 0,8 0,6 0,4 0,2 0,0 0

2000

4000

6000

8000

Time (sec) Figure 4. Cl2 activation study. Experimental conditions: 35 Cl2 pulses (230 s pulse length, 1.8 µmol Cl2 / pulse), He flow: 10 ml/min, 10 mg of polymer A synthesized from 1,4dicyanobenzene.

The irreversible addition and substitution of Cl2 is hypothesized to occur primarily on the most reactive sites, i.e., defect sites and unsaturated carbons located at zig-zag edges and terminating positions.48 The amount of irreversible Cl2 uptake and of HCl formed on the investigated carbon materials during Cl2 treatment at 473 K are summarized in Table 4. Table 4. Cl2 adsorption on N-containing carbon materials at 473 K.

Polymer

Sites for irreversible Defects / adsorption/ edge sites aromatic ring

Irreversible Cl2 uptake (mmol/g)

N-function A1

A

2

0.45

S2 0.21

Accessible C- sites A

S

1.09

1.00

HCl formation (mmol/g)

R-NH2

0.21

Accessible C sites

1.00


16

Page 17 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

B

C

D

P

R

0

0.73

0.29

0

0

0.29

0

4

0.14

0.08

4.23

2.11

0.084

2.11

4

0.12

0.17

4.13

1.78

0.17

1.78

P = polymeric network as proposed in Figure 1 and Figure S2, R = N and O-based substituents. 1 A= Addition determined based on the XPS and MS analysis results, 2S = Substitution reaction. Additional information for the calculated values are provided in Table S7.

The results show that the total Cl2 uptake does not correlate with the specific surface area of the catalysts. The irreversible Cl2 uptake observed at the beginning is strictly proportional to the concentration of HCl formed and it is always lower than the initial H content of the materials (28 - 30 mmol/g for polymers A, C and D and 18 mmol/g for polymer B) indicating that only C-H bonds at defects (indicated in Table 4) may be converted by Cl2 chemisorption, while other H containing sites and functional groups remain at least partly unaffected. Reactive terminal functional groups of the polymers such as –C≡N or –C-NH2 groups (Table 1) are activating the aromatic rings by increasing the electron density of the π-system (higher nucleophilicity) and should favor the facile and irreversible oxidative addition and substitution reactions with Cl2.49 To experimentally test this hypothesis, 1,4-dicyanobenzene was used as a probe catalyst (experimental measurement details summarized in supporting information Figure S11). As expected, the Cl2 pulses conducted in the presence of aromatic nitriles resulted in irreversible Cl2 addition and HCl formation.


17

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 39

In separate control experiments, also the role of the aromatic amines in the Cl2 activation has been explored. 1,4-Diaminobenzene, for example, reacted readily with Cl2, results in the immediate formation of chlorination products (Figure S15). Overall, aromatic amines promote the irreversible consumption of Cl2 via secondary reaction pathways.52,53 Chlorination of aromatic amines does not need highly polar solvents or an added catalyst, frequently required for other electrophilic aromatic substitutions.52 The formation of a Nchloramine as a reaction intermediate facilitates the electrophilic Clδ+ via the heterolytic cleavage of N-Cl bond, able to further promote the substitution of H for Cl. In the current experiments, the initial Cl2 uptake and the number of free accessible aromatic sp2 carbon sites at vicinal ortho and para positions to an amine or nitrile functionality were directly correlated (Figure 5 (a)).

Figure 5. Correlation between free accessible sp2 C sites in the aromatic ring in the terminal polymer structure and initial Cl2 uptake at 473 K (a) and amount of HCl formed at 473 K (b). Calculation details are presented in Table 5 and S6.

In case of polymer B (all active C sites blocked by Cl2) the irreversible Cl2 uptake and HCl is enabled only on R-NH2 sites (Table 1) leading to the formation of chloramines and HCl as specified in the general Equ. 3.60


18

Page 19 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

The fact that the Cl2 uptake was 3.5 times higher than the HCl release indicates that a noncovalent Cl2 pre-reactive complex is formed with the N-atoms in triazine, as suggested by DFT calculations and proposed in Scheme 3.55 The Cl2 uptake on the amine functionalities and triazine units in polymer A, C and D are listed in Table 4. The linear dependence of the irreversible Cl2 uptake from the concentration of nucleophilic carbon at ortho and para positions of the aromatic rings in polymer A, C and D suggests that these sites favor electrophilic Cl2 adsorption and formation of strong C-Cl covalent bonds (Scheme 1).50,51,52

Scheme 1. Proposed reaction sequence that would account for the aromatic amine promoted irreversible Cl2 interaction and HCl formation.

Compared to polymer A, Cl2 addition to polymer C and D results in a four times higher Cl2 uptake and two times higher HCl formation, because the aromatic linkers used in both materials have a larger fraction of accessible reaction sites (Figure 5 (a) and (b)). Quantitative analysis of the irreversible Cl2 consumption via addition and substitution reactions with polymer C and D showed that about two times more Cl2 (4.37 mM Cl2 / g for polymer C and 4.26 mM Cl2 / g for polymer D) were consumed than in substitution reactions (2.19 mM Cl2 in polymer C and 1.95 mM Cl2 in polymer D) (Table 4 and Figure 5).


19

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

*C + Cl2

*C

Page 20 of 39

δ− δ+ Cl Cl

(activated Cl2)

Scheme 2. Schematic representation of the transient formation of activated Cl2. After the saturation of these most reactive sites, Cl2 interacted reversibly with the electropositive C atoms of the deactivated aromatic ring (Scheme 2).

3.2.2 Raman analysis of the catalytic reaction The Cl2 adsorption on N-containing material (polymer A) synthesized from 1,4dicyanobenzene was further studied using in situ Raman spectroscopy (Figure 6).

Figure 6. In situ Raman spectroscopic study of Cl2 activation on triazine based polymer A (for structural information see Figure 1). In (a) is shown the irreversible addition and substitution reactions primarily taking place in the first cycle of Cl2 addition. In (b) the reversible interaction during the second cycle is presented.

The increase in both D- and G- Raman band intensity is attributed to the Cl2 adsorption on the active sites of amorphous and graphitic phases. In the first cycle, the irreversible addition and substitution reactions on the more active sites of the carbon material will lead to irreversible changes in the Raman spectra of the mesoporous carbon (Figure 6 (a)). Following the saturation of these sites, the reversible increase and decrease of G and D band intensity


20

Page 21 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

demonstrates the reversible adsorption and desorption of Cl2 (Figure 6 (b)). On planar HSAG (Figure S12) Cl2 exposure beyond the irreversible addition and substitution, did not lead to reversible adsorption of Cl2 (identical spectra in presence and absence of Cl2). These results strongly support the hypothesis that, the strong electronic influence of inserted pyridinic and quaternary N is responsible for the formation of active carbon sites at which the reversible molecular Cl2 adsorption is taking place. The electronic nature of the surface activated Cl2 under steady state conditions was probed by in situ ESR spectroscopy. Transient formation of radical Cl2 species on N-containing C has been excluded on the basis of the unchanged signal in the ESR spectra obtained after the Cl2 exposure (Figure S13). 3.2.3 Probing the nature of the transiently formed activated chlorine To further discriminate between ionic and radical activation routes for Cl2 the chlorination of CH4 was used as probe reaction. The substitution of H with Cl proceeds solely via a radical mechanism.54 The absence of substitution reactions induced by the carbon of methane (Figure 7) indicates that the 1,4-dicyanobenzene derived carbon did not activate Cl2 by a radical pathway.


21

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 39

Figure 7. Probing activated Cl2 with CH4. Experimental conditions: 100 pulses (first 50 Cl2 pulses till steady state conditions, 230 s pulse length, 1.8 µmol Cl2 / pulse), 5 vol. % CH4 in He flow: 10 ml/min, T = 473 K, 10 mg of mesoporous carbon A (shown in Figure 1).

Similarly, the Cl2 activation on the N-containing carbons C and D obtained from 9,10anthracenedicarbonitrile and 4,4`-biphenyldicarbonitrile using Raman and ESR spectroscopy, confirmed that the reversibly adsorbed Cl2 on the surface has an ionic character as depicted in Scheme 2. The Cl2 activation on carbon materials with a high content of amine and pyrrolic-N functionalities mainly promoted the aromatic *C-H bond substitution by Cl2 and led to the formation of covalently bound –Cl on sp2 carbon site. In the corresponding Raman spectra (Figure 8), a more accentuated irreversible shift of about 41 cm-1 (from 1584 cm-1 black spectra to1543 cm-1 in the orange spectra) and intensity decrease in the G- band points to a decrease of the well-ordered graphitic phase and to a concomitant relative increase in disordered carbon during the irreversible chlorination. As indicated by the decrease of the Dband intensity, the chlorination of the amorphous phase due to the facile adsorption of Cl2 resulted in an overall consumption of the less-ordered amorphous phase and to the formation


22

Page 23 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

of volatile chlorinated short chain hydrocarbons (formation of CH2Cl2, CHCl3, CCl4, C6Cl6 as shown in in situ MS measurements).

Figure 8. In situ Raman spectra of activated Cl2 on material B obtained from tetrachloroterephthalonitrile. Irreversible interactions observed during the first cycle are shown in (a) while during the second cycle of repetitive Cl2 additions mainly reversible interactions were observed (b).

The material B derived from tetrachloroterephthalonitrile includes sp2-C hindered by the covalently bound –Cl and results in a different type of Cl2 interaction. It is hypothesized that the site for the reversible Cl2 interaction with this catalyst is the lone electron pair of the Natom in a triazine molecule resulting in the formation of a weak physisorbed non-covalent adduct (Figure S20), as depicted in Scheme 3. This adduct does not lead to product formation and is in line with previous DFT calculations of Cl2 interactions on triazine.55

N

N N N

+

N

Cl2

Cl

Cl

N

Scheme 3. Proposed pre-reactive complex of 1,3,5-triazine and Cl2.


23

ACS Catalysis

3.2.4 In situ Raman analysis of Cl2 activation in the presence of N-containing model molecules To explore Cl2 activation at the molecular level, the interactions with N-containing model molecules were examined using in situ Raman spectroscopy. A short overview of the Cl2 interaction ability of model carbon molecules is compiled in Table 5.

Table 5. Summary of Cl2 activation study on well-defined N-containing model molecules using in situ Raman spectroscopy. N-type Selected model molecules

Type-1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 39

Cl2 activation Adsorption

Py.-N (at. %)

Quart.-N (at. %)

9.03

-

160, 274, 398

743

907

Reversible

9.09

-

-

-

-

Weak physisorption

Raman bands (cm-1)


24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Type-2

Page 25 of 39

8.34

-

-

-

-

No interaction

39.85

-

-

-

-

Weak physisorption

33.34

-

-

-

-

Weak physisorption

-

4.08

307, 352, 399

-

-

Reversible

-

4.25

272

-

-

Reversible

-

6.66

-

-

-

No interaction

The role of pyridinic-N was studied by considering two model compounds, i.e., pyrazine and s-triazine. The band at 1013 cm-1 corresponds to a pyrazine ring vibration and the band around 3000 cm-1 is due to C-H bond vibration (supporting information, Figure S16). Cl2 addition to the pyrazine increased the intensities of Raman bands and further resulted in the irreversible adsorption of activated Cl2 species. Irreversible chlorination converted the solid pyrazine to liquid mono and dichloro pyrazine (liquid state 1H and 13C-NMR, Figure S17). As a consequence, new bands appeared at 678 cm-1 and 1224 cm-1 (supporting information, Figure S16) in the Raman spectra, characteristic of the new covalent C-Cl bonds.56 In the case of triazine, the Cl2 interaction caused primarily an increase in the Raman band intensity due to the activation of Cl2 on the electron deficient carbon sites as depicted in Scheme 4.

Scheme 4. Proposed Cl2 interaction with 1,3,5-triazine.


25

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 39

In this case Cl2 is activated on the carbon site next to the pyridinic N-atom by the formation of a *C····Cl2 adduct of ionic character. In the presence of the transiently generated ionic Cl2 a more accentuated decrease in Raman bands intensity was caused by the change in the bond arrangement (formation of new C-Cl covalent bonds) and realignment of the triazine molecules as the phase transition is approached (supporting information, Figure S18).58 Melamine and cyanuric chloride have all carbon atoms adjacent to the pyridinic nitrogen, functionalized by electron donating –NH2 and electron withdrawing –Cl, respectively. After admission of Cl2 to melamine and cyanuric chloride, changes in the Raman spectra were not observed, because the sites allowing the activation and reaction of Cl2 were blocked (Figures S19 and S20). These experiments indicate that the site for Cl2 activation is an electron deficient C-atom and not the pyridinic N-atom. Pyridinic N-modification only enhances the electron localization and creates a new site on sp2-C of electrophilic character for the Cl2 activation (Scheme 3 and 4). The requirement for a site in N-containing carbon for the reversible Cl2 activation was explored by selection of three further model materials; 2,6-dichloropyridine, 2,4,6trichoropyridine and 2,6-dichloropyridine N-oxide. Cl2 addition on 2,6-dichlropyridine (Figure S21 (a)) resulted initially in an increase in the Raman band intensities with exposure time. However, exposure to Cl2 (> 7 min continuous Cl2 flow) decreased the intensity of these bands, while in parallel bands at 160, 274, 398, 743, 884 and 907 cm-1 appeared (Table 5 and Figure S21 (b)). The more intense band at 398 cm-1 is attributed to the symmetric stretching vibration of the polarized Cl2, the bands at lower wavenumbers to the bending mode of the activated Cl2. The asymmetric vibration modes of the polarized Cl2 are observed at higher wavenumbers, at 884 and 907 cm-1. Combining these observations and attributions


26

Page 27 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

allows us to propose the Cl2 activation mechanism on nitrogen containing moieties as depicted in Scheme 5.

Scheme 5. Proposed Cl2 activation mechanism on resonance stabilized 2,6-dichloropyridine. It is hypothesized that Cl2 interacts with the positively charged carbon in the para-position of 2,6-dichloropyridine, forming a *C····Cl2 adduct. This hypothesis was confirmed by the inability of 2,4,6-trichloropyridine to adsorb and activate Cl2. The Raman spectra during Cl2 addition showed only a weak reversible interaction without formation of new bands (Figure S22). Similarly, 2,6-dichlropyridine N-oxide (absence of a positive charge at the C atom in para position to N) during Cl2 treatment does not lead to new Raman bands (Figure S23). Overall, the experiments support the hypothesis that an accessible electron deficient carbon must be generated to adsorb and activate Cl2. In order to understand the role of the second form of N-modification (Table 1), the quaternary N-atoms in carbon, 1,3-bis(2,4,6-trimethylphenyl)-1,3-dihydro-2H-imidazol-2ylidene (BTDIY, shown Figure 3 (a)) was selected. Upon Cl2 addition to BTDIY new Raman


27

ACS Catalysis

bands appeared reversibly at 272 cm-1 and were attributed to the formation of the [BTDIY····Cl2] adduct (Figure 9, Scheme S1).

Normalized intensity (a.u.)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 39

a 1,0 0,8

b c d

N2 flow

272 BTDIY + Cl2

5 min Cl2 20 min Cl2

b~c

30 min N2

0,6

d

0,4 0,2

BTDIY

a

0,0 200

225

250

275

300

325

350

Raman shift (cm-1) Figure 9. In situ Raman spectroscopy of Cl2 interaction with BTDIY. The in situ ESR spectrum (recorded at 313 K, 101 kPa and Ar) showed a strong signal at a g value 1.9882 ± 0.0012, confirming the presence of quaternary-N stabilized paramagnetic sites on BTDIY (Figure S14). The two quaternary N atoms stabilize the carbene through σelectron acceptor and π-electron donor interaction.24 Changing the gas from Ar to Cl2 diminished the ESR signal intensity, indicating the ionic character of the [BTDIY····Cl2] adduct (Figure S14 and Scheme S1). In order to test this attribution, we also explored the interactions of 1,3-bis(2,4,6trimethylphenyl)imidazolium chloride (BTIC, Figure 3b). Cl2 addition resulted in a new band at 352 cm-1, attributed to Cl2 on BTIC (see SI, Figure S24). The broad band at 352 cm-1 is the combination of three Raman bands (two strong bands at 307, 352 and one weak at 399 cm-1, Figure 10).


28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Normalized Intensity (a.u.)

Page 29 of 39

1,0 0,8 0,6

Pre peak Peak fitting Peak 1 Peak 2 Peak 3 Baseline

352

307 399

0,4 0,2 0,0 250

300

350

400

450

-1

Raman shift (cm ) Figure 10. Deconvolution of Raman spectra of activated Cl2 on BTIC. Purging with N2 led to the disappearance of all bands, indicating the reversibility of the adduct formation. The three Raman bands suggests the formation of the trichloride monoanion (Cl3-).58 The band at 352 cm-1 is attributed to the polarized Cl2, while the weak Raman bands at 307 and 399 cm-1 are assigned to Cl−···Cl2 interactions. The bound Cl2 with ionic character was indirectly confirmed by in situ ESR spectroscopy, as it failed to show signals of Cl2 with radical character (supporting information, Figure S25). The critical importance of the electrophilic C atom between the quaternary N atoms (Scheme

S1)

was

tested

by

exposing

1-ethyl-2,3-dimethylimidazolium-

trifluoromethanesulfonate (EDITM, Figure 3c), which contains a methylene group, in between two quaternary N-atoms (supporting information, Figure S26). The absence of new Raman bands upon exposure to Cl2 allowed us to conclude that a non-substituted electrophilic C atom is required for an active site that is able to activate Cl2 via polarization.

3.3 Mechanistic study for COCl2 synthesis on N-containing carbon


29

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 39

Previous work on fullerene-type model catalysts has shown that COCl2 synthesis proceeds by the reaction of CO with adsorbed Cl2.25 To better understand this stepwise chemistry, the interaction of CO with a Cl2 saturated N-containing carbon material from 1,4dicyanobenezene (polymer A) was studied by in situ Raman spectroscopy (Figure S27). The results show hardly any impact of CO at 473 K on the sample saturated with Cl2. Under the same operating conditions, however, pulses consisting of Cl2 in 5 vol. % CO led to a gradually increasing concentration of COCl2 on this N-containing carbon material (Figure 11).

Figure 11. (a) Cl2 adsorption at 473 K in 5 vol. % CO and (b) COCl2 synthesis. Experimental conditions: 35 Cl2 pulses (230 s pulse length, 1.8 µmol Cl2 / pulse), 5 vol. % CO in He flow: 10 ml/min, 10 mg of N-containing carbon material (polymer A).

At the start, the Cl2 pulses resulted in HCl (1.17 mmol/g) formation in concentrations similar to pulsing Cl2 in He (1.2 mmol/g, Figures 6 and Table 4). This shows that the most reactive carbon sites interact selectively with Cl2 without affecting the CO interactions. Following the (partial) saturation of these sites after 1000 s, Cl2 was detected in gas phase. At the same time, COCl2 was also detected by gas phase. Thus, we conclude that only after a significant fraction of the most reactive sites on carbon have been saturated (i.e., after


30

Page 31 of 39

irreversible Cl2 addition and substitution reactions) sufficient probability exists that Cl2 is reversibly activated on the electrophilic carbon sites of the N-containing materials. The reversibly adsorbed activated Cl2 is assumed to be the key species to react with CO resulting in COCl2 formation, as demonstrated in ref. [25]. The formation of this critical reversibly bound activated complex strongly depends upon the Cl2 reactions with carbon surfaces

10 pulse (A s) / m2 x 10-10

bearing active carbon sites in various electronic and steric environment.25

COCl2 ion current x time per

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

10

Polymer D

Polymer A

8 6

Polymer C 4 2

Polymer B

0

Polymers Figure 12. Comparisons of phosgene formation on N-containing carbon catalysts. Experimental conditions: 50-100 Cl2 pulses (230 s pulse length, 1.8 µmol Cl2 / pulse), 5 vol. % CO in He flow: 10 ml/min, 10 mg of N-containing carbon material. COCl2 estimated after steady-state is achieved.

To explore the impact of electronic and steric variations induced by the N-modification on COCl2 synthesis, catalysis was conducted with the series of N-modified carbon materials (Figure

12).

The

materials

synthesized

from

1,4-dicyanobenzene

and

4,4`-

biphenyldicarbonitrile catalyzed COCl2 formation to a similar extent (0.10 µmol/m2 × s and 0.12

µmol/m2 ×

s,

respectively).

The

carbon

material

synthesized

from

the

tetrachloroterephthalonitrile precursor, in contrast, hardly showed catalytic activity. Note that this material has primarily carbon sites on the aromatic ring blocked with covalently bound –


31

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 39

Cl, demonstrating that the covalent C-Cl bonds at the potential active carbon sites block reactivity completely. All studied N-modified carbon materials were highly stable (Figure 11 (b), SI Figure S28, S29). Important kinetic results are summarized in Table 6. Table 6. Kinetic analysis results of the carbon catalyzed COCl2 formation a COCl2 Heat of Cl2 b formation rate at Eaapp adsorption Catalysts 473 K [kJ/mol] [kJ/mol]c 2 [µmol/s m ] 33d Polymer A 0.10 34

Reaction order

1

Activated carbon

0.036

56

43d

1

C60 fullerene

0.120

18

10

1

a

All catalysts were activated in He (10 cm3/min) for 1 h at 423 K. ± 2 kJ/mol. All the catalytic reactions were performed in a temperature range of 333-623 K. c ± 3 kJ/mol. All the measurements were carried out in the presence of 50 – 200 mg of carbon by using 0.2 – 2.7 µmol Cl2 diluted in He atm., ambient temperature. The heat of adsorption was determined for the reversible Cl2 adsorption after the sites where irreversible processes are enabled were completely saturated. d In the case of irreversible Cl2 adsorption on the most reactive sites of an activated carbon, 134 kJ/mol (for activated carbon) and 141 kJ/mol (for polymer A) were detected. b

It is most significant to note that for the associative Cl2 adsorption and activation on a sp2hybridized carbon site is strongly influenced by the hybridization and the local electronic environment of the carbon. Adsorption measurements for Cl2 on various structurally different carbons (shown in Table 6) showed a weak adsorption on a mixed sp2/sp3-hybridized carbon site of a C60 fullerene (10 kJ/mol, Table 6 and Ref. 25) while on a pure sp2-hybridized carbon, the higher heats of adsorption (43 kJ/mol) indicate a better accommodation and a more localized adsorption of Cl2. Note, that the radical pathway observed in the case of C60 catalyzed COCl2 synthesis leads to lower energy of activation (Table 6).

25

The identical

overall rates (Polymer A and C60 fullerene) points to a higher apparent transition entropy. It is


32

Page 33 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

hypothesized, that this higher entropy is associated with a higher concentration of active carbon sites in Polymer A compared to C60 fullerene. Nitrogen insertion in the aromatic ring structures of carbon materials (see Figure 1) results in the electronic and steric influence (structural constraint caused by the pyramidalization of the nitrogen63) on the aromatic sp2hybridized carbon. This in turn leads to the formation of new electron-deficient carbon sites for the molecular Cl2 adsorption and polarization that present an energetically more favorable reaction pathway (Eapp = 34 kJ/mol compared to Eapp = 56 kJ/mol) relative to activated carbons. The Cl2 activation on electropositive carbon sites of a nitrogen modified surface is responsible for the generation of activated molecular chlorine intermediate with ionic character able to react with CO (Figure 9, 10, 11), i.e., leading to an ionic pathway to phosgene. The COCl2 formation on this ionic pathway is also the reason for the outstanding material stability of the nitrogen containing carbon catalysts (Figure 11). The transient generation of reactive intermediates with radical nature could be ruled out in this case, because of the absence of signals during in situ ESR experiments (Figure S13 S14 and S24) and by using CH4 as a probe molecule for the radical substitution reaction (Figure 7). Thus, combining spectroscopic analyses and reaction for the nitrogen substituted carbon catalysts we propose the catalytic cycle depicted in Scheme 6.


33

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 39

Scheme 6. Proposed reaction sequence that counts for COCl2 formation on N-containing carbon materials.

The exposure of nitrogen functionalized carbon materials to Cl2 results in associative adsorption of Cl2 on the electron deficient active carbon sites and the concomitant formation of an intermediate activated ionic Cl2. The subsequent attack of physisorbed CO is hypothesized to lead to the formation of acylchloride cation [+δC(O)Cl] and weakly bound Clδ

in the transition state that decomposes with the concerted reaction of reactive species

[+δCOCl] with Cl-δ to yield COCl2. This mechanism is consistent with the measured reaction orders of Cl2 and CO, both of which are 1 (Table 6). It pictures the reversible polarized Cl2 as the intermediate that can be detected and characterized by spectroscopy. However, CO neither perturbs the polarized Cl2 detectably nor has a measurable uptake of adsorption. CO is only weakly physically adsorbed and there exists a barrier of 34 kJ/mol to form the activated complex with the polarized Cl2. As the intermediate does not accumulate on the carbon catalysts we hypothesize that the reaction proceeds with a concerted decomposition of the activated complex to COCl2.

CONCLUSIONS Combining spectroscopic results and reaction data it was shown that the N modification of C materials generates new active sites for the catalyzed reaction of Cl2 and CO to COCl2. Exploring Cl2 activation on model carbon materials with different degrees of functionalization showed that the energetically favored irreversible addition and substitution reactions of Cl2 at the defect sites and terminating carbon atoms (mainly located in the amorphous bulk part of the material) occur only in the initial stages of the reaction. These reactions lead to a high initial Cl2 consumption and the formation of HCl and CH4-nCln.


34

Page 35 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

The insertion of nitrogen substituents with electron withdrawing properties into the wellorganized graphitic carbon layers enables the formation and stabilization of electropositive carbon sites with enhanced adsorption ability for Cl2. In contrast to pure sp2/sp3 mixed hybridized carbon surfaces, where the localized Cl2 adsorption results in the formation of key intermediates with radical character, the associative adsorption of Cl2 on the electropositive carbon sites in the vicinity of nitrogen leads to polarized Cl2. In presence of CO, COCl2 was formed with very high selectivity. The ability to selectively generate these ionic intermediates during the COCl2 synthesis has been identified as the main reason for the long-term stability of the N-functionalized carbons.

ASSOCIATED CONTENT Supporting Information Available: The material includes experimental details, additional spectra related to in-depth materials characterization and in situ spectroscopic studies. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author * J. A. Lercher: E-mail: [email protected], Tel.: (+)49-89-28913540. Fax: (+)4989-28913544. E. E. Ember: E-mail: [email protected], Tel.: (+)49-89-28913545. Fax: (+)49-8928913544. Notes The authors declare no competing financial interest.


35

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 39

ACKNOWLEDGMENT The authors acknowledge support from European Community's Seventh Framework Program [FP7/2007-2013] under grant agreement no. NMP-LA-2010-245988 (INCAS). The authors also thank Dr. Stefan Roggan, Prof. Mleczko and Jens Sicking from Bayer Technology Services GmbH for the fruitful discussions and XPS analysis of the carbon materials. We would also like to thank Xaver Hecht for N2 physisorption experiments, Martin Neukamm and Sebastian Foraita for HR-SEM measurements. REFERENCES (1) Serp, P.; Figueiredo, J. L. Carbon Materials for Catalysis; John Wiley & Sons: Hoboken, NJ, 2009. (2) Trogadas, P.; Fuller, T. F.; Strasser, P. Carbon 2014, 75, 5–42. (3) Titirici, M. M.; White, R. J.; Brun, N., Budarin, V. L.; Su, D. S.; Monte, F. d.; Clark, J. H.; MacLachlan, M. J. Chem. Soc. Rev. 2015, 44, 250–290. (4) Burchell, T.D. Carbon materials for advanced technologies, Pergamon Press, 1999. (5) Wickramaratne, N. P.; Jaronie M. ACS Appl. Mater. Interfaces 2013, 5, 1849–1855. (6) Bégina, D.; Ulrichb, G.; Amadoua, J,; Su D. S.; Pham-Huua, C.; Ziessel, R. J. Mol. Catal. A: Chem. 2009, 302, 119–123. (7) Puri, B.R., Bansal, R.C., Carbon 1966, 3, 523–539. (8) Zhang, P.; Gong Y.; Li, H.; Chen Z.; Wang, Y., Nat. Commun. 2013, 4, 1593. (9) Zheng, F.; Yang, Y.; Chen, Q. Nat. Commun. 2014, 5, 5261. (10) Zhang, J.; Liu, X.; Blume, R.; Zhang, Z.; Schlogl, R.; Su, D. Science 2008, 322, 73–77. (11) Yang, L.; Jiang, Shujuan.; Zhao Y.; Zhu, L.; Chen, S.; Wang, X.; Wu, Q.; Ma, J.; Ma, Y.; Hu, Z. Angew. Chem. Int. Ed. 2011, 50, 7132–7135. (12) Chizari, K.; Deneuve, A.; Ersen, O.; Florea, I.; Liu, Y.; Edouard, D.; Janowska, I.; Begin D.; Pham-Huu, C. ChemSusChem 2012, 5, 102–108. (13) Gong, K. P.; Du, F.; Xia Z. H.; Durstock, M.; Dai, L. M. Science 2009, 323, 760–764. (14) Tang, Y.; Allen, B. L.; Kauffman, D. R.; Star A. J. Am. Chem. Soc., 2009, 131, 13200– 13201. (15) Li, Q.; Zhang, S.; Dai, L.; Li, L. S. J. Am. Chem. Soc., 2012, 134, 18932–18935. (16) Ai, K,; Liu, Y.; Ruan, C.; Lu, L.; Lu, G. M. Adv. Mater. 2013, 25, 998–1003.


36

Page 37 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

(17) Zhang, L.; Xia, Z. J. Phys. Chem. C, 2011, 115, 11170–11176. (18) Gong, K. P.; Du, F.; Xia, Z. H.; Durstock, M.; Dai, L. M. Science 2009, 323, 760–764. (19) Chung, H. T.; Won, J. H.; Zelenay, P. Nat. Commun. 2013, 4, 1922. (20) Yu, D.; Nagelli, E.; Du, F.; Dai, L. J. Phys. Chem. Lett. 2010, 1, 2165–2173. (21) (a) Chizari, K.; Deneuve, A.; Ersen, O.; Florea, I.; Liu, Y.; Edouard, D.; Janowska, I.; Begin D.; Pham-Huu, C. ChemSusChem 2012, 5, 102–108. (b) Xu, X.; Li, Y.; Gong, Y.; Zhang, P.; Li, H.; Wang, Y. J. Am. Chem. Soc. 2012, 134, 16987–16990. (c) Li, M.; Xu, F.; Li, H.; Wang, Y. Chem. Sci. Technol. 2016, 6, 3670 – 3693. (22) Yu, D.; Xue, Y.; Dai, L. J. Phys. Chem. Lett. 2012, 3, 2863−2870. (23) Baltazzi, E.; Krimen, L. I. Chem. Rev. 1963, 63, 511−556. (24) Herrmann, W. A.; Kocher, C. Angew. Chem. Int. Ed. 1997, 36, 2162–2187. (25) (a) Gupta N. K.; Pashigreva A. V.; Pidko E. A.; Hensen E. J. M.; Mleczko L.; Roggan S.; Ember. E. E.; Lercher J. A. Angew. Chem. Int. Ed. 2015, 54, 1–6. (b) Centi G., Barbera K., Perathoner S., Gupta N. K., Ember E. E., Lercher J. A. ChemCatChem 2015, 7, 3036– 3046. (26) Khachatryan L.; Dellinger B. Chemosphere 2003, 52, 709–716. (27) Kuhn, P.; Thomas, A.; Antonietti, M. Macromolecules 2009, 42, 319–326. (28) American Conference of Governmental Industrial Hygienists: “Phosgene”, Documentation of the Threshold Limit Values for Substances in Workroom Air (3rd ed., 2nd printing), Cincinnati, 1974. (29) Barrett, E. P.; Joyner, L.G.; Halenda, P.P. J. Am. Chem. Soc., 1951, 73, 373–380. (30) Gregg, S. J., Kenneth, S. W. S. Adsorption Surface Area and Porosity ed. 2, Academic Press Inc., U.S., London, 1982. (31) Lippens, B.C.; Deboer, J.H. J. Catal. 1965, 4, 319–323. (32) NIST X-ray Photoelectron Spectroscopy Database, Database 20, Version 3.5, http://srdata.nist.gov/xps/Default.aspx

(33) Ferrari, A. C.; Robertson, J.; Phil. Trans. R. Soc. Lond. A 2004, 362, 2477–2512. (34) Sadezky, A; Muckenhuber, H.; Grothe, H.; Niessner, R.; Pöschl, U. Carbon, 2005, 43, 1731–1742. (35) Haghseresht, F.; Lu, G. Q.; Whittaker, A. K. Carbon, 1999, 37, 1491–1497. (36) Reich, S.; Thomsen, C.; Phil. Trans. R. Soc. Lond. A 2004, 362, 2271–2288. (37) Lin, Y.; Pan, X.; Qi, W.; Zhang, B.; Su, D. S. J. Mater. Chem. A 2014, 2, 12475–12483. (38) Jagst, E., PhD thesis, Freien Universität Berlin, Germany 2010. (39) Shen, W.; Fan, W. J. Mater. Chem. A 2013, 1, 999–1013.


37

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 39

(40) Luo, Z.; Lim, S.; Tian, Z.; Shang, J.; Lai, L.; MacDonald, B.; Fu, C.; Shen, Z.; Yu, T.; Lin, J. J. Mater. Chem. 2011, 21, 8038–8044. (41) Wang, S.; Yu, D.; Dai, L. J. Am. Chem. Soc. 2011, 133, 5182–5185. (42) Retzko, I.; Unger, W. E. S. Adv. Eng. Mater. 2003, 5, 519–522. (43) Schmitz C. H.; Rang C.; Bai Y.; Kossev I.; Ikonomov J.; Su Y.; Kotsis K.; Soubatch S.; Neucheva O.; Tautz S.; Neese I. S.; Steinrück H.-P.; Gottfried J. M.; Dötz K. H.; Sokolowski M., J. Phys. Chem. C 2009, 113, 6014–6021. (44) Jagst, E., PhD thesis, Freien Universität Berlin, Germany 2010. (45) Yanga, D.; Velamakannia, A.; Bozoklub, G.; Parka, S.; Stollera, M.; Pinera R. D.; Stankovichc, S.; Junga, I.; Fieldd D. A.; Ventrice C. A.; Ruoffa, R. S. Carbon 2009, 47, 145– 152. (46) Wilgus, C. P.; Downing, S.; Molitor, E.; Bains, S.; Pagni, R. M.; Kabalka, G. W. Tetrahedron Lett. 1995, 36, 3469–3472. (47) Krieble, V. K.; Noll, C. I. J. Am. Chem. Soc. 1939, 61, 560–563. (48) Chen, J. P.; Yang, R.T. Surf. Sci. 1989, 216, 481–488. (49) (a) Watts L. US3816505 1974. (b) Baciocchi, E.; Cacace, F.; Ciranni, G; Illuminati, G. J. Am. Chem. Soc., 1972, 94, 7030–7034. (50) Legon, A. C.; Thumwood, J. M. A. Phys. Chem. Chem. Phys. 2001, 3, 1397–1402. (51) Olah G. A. Angew. Chem. Int. Ed. 1973, 12, 173–212. (52) O'Connell, J. L.; Simpson, J. S.; Dumanski, P. G.; Simpson, G. W.; Easton, C. J. Org. Biomol. Chem. 2006, 4, 2716–2723. (53) Gassman, P. G.; Campbell, G.A. J. Am. Chem. Soc. 1971, 93, 2567–2569. (54) (a) Dayid B., US2839589 A, 1958. (b) Cass O. W., US2406195 A, 1946. (55) Akićand, D. S.; Vrcek, V. J. Phys. Chem. A 2012, 116, 1298−1306. (56) Bernussi, A. A.; Gualberto, G. M. J. Raman Spectrosc. 1987, 18, 93–95. (57) Grundmann, C.; Kreutzber, A. J. Am. Chem. Soc., 1955, 77, 44–48. (58) Arduengo, A. J.; Davidson, F.; Dias, H. V. R.; Goerlich, J. R.; Khasnis, D.; Marshall, W. J.; Prakasha, T. K. J. Am. Chem. Soc. 1997, 119, 12742–12749. (59) Kuhn, N.; Fahl, J.; Fawzi, R.; Maichle-Mößmer, C.; Steimann, M. Z. Naturforsch. B Chem. 2014, 53, 720–726. (60) Brückner, R.; Haller, H.; Ellwanger, M.; Riedel, S. Chem. Eur. J. 2012, 18, 5741–5747. (61) Abrams, L.; Cicha, W. V.; Manzer, L. E.; Subramoney, S., Stud. Surf. Sci. Catal., 2000, 130, 455–460.


38

Page 39 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

(62) Jenkins, R. L.; Haskins, J. E.; Carmona, L. G.; Baird, R. B.; Arch. Environm. Contain. Toxicol., 1978, 7, 301–315. (63) Waluk, J.; Methods in Stereochemical Analysis Conformational Analysis of Molecules in excited states, Wiley & Sons Ltd, New York, 2000.

Insert Table of Contents Graphic and Synopsis Here


39

Nitrogen Modified Carbon Nano-Materials as ... - ACS Publications

Recommend Documents