A Computational Exploration of the CO Adsorption in Cation

Subscriber access provided by Mississippi State University Libraries

Comment

Comment on “A Computational Exploration of the CO Adsorption in Cation-Exchanged Faujasites” (J. Phys. Chem. C 2012, 116, 11195) Olivier Cairon J. Phys. Chem. C, Just Accepted Manuscript • Publication Date (Web): 25 Sep 2013 Downloaded from http://pubs.acs.org on September 26, 2013

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.

The Journal of Physical Chemistry C 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 7

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

The Journal of Physical Chemistry

Comment on “A Computational Exploration of the CO Adsorption in Cation-Exchanged Faujasites” O. Cairon* Chem. Des Pyrénées - 64160 Serres-Morlaàs, France E-mail: [email protected]

In a recent paper by Nour, Berthomieu, Yang and Maurin, molecular simulations were employed to rationalize CO adsorption on NaY and NaX Faujasites.1 Nour et al. compared results obtained from Grand Canonical Monte Carlo (GCMC) simulations to some experimental data in the literature, especially data from FTIR analyses.2-4 Their interpretation of the FTIR literature seems somewhat erroneous and could mislead the reader regarding CO adsorption on NaY and NaX zeolites. Moreover, recent FTIR experimental results as well as other results combining theoretical approaches that deal specifically with CO adsorption on NaY and NaX zeolites have been ignored.5-10 By taking the time to clarify the issue and better situate Nour et al.’s contribution in relation to the literature, we will give the reader for the additional information needed to make a valuable comparison between FTIR experiments and GCMC simulations. Here, we chronologically review the major points of discussion in connection with the literature of CO adsorption on NaY and NaX zeolites. The conclusions of Nour et al. are not supported by these previous experimental FTIR studies or others combining theoretical molecular approaches.

The first point we would like to discuss is the microscopic adsorption behavior of NaY and NaX. To start with, let us briefly remember the particular distribution of cations within the two NaY and NaX zeolites. As rightly noted by Nour et al., NaY and NaX have the same type of framework but due to their respective Si/Al ratios, these two zeolites differ in both the number of Na+ cations they contain and the Na+-distribution over the different cationic sites. Taking into account that CO only probes the Na cations sitting in the supercage (s.c.), this means only Na+ cations occupying either SII or SIII’ and SIII will interact with CO. In addition, it’s well known that for the NaY (deshydrated form), neither SIII’ nor SIII are populated but only SII with 4 Na+ per s.c. (i.e., 32 Na in SII per unit cell, u.c.). For the NaX (per u.c), 32 Na+ occupy the SII and 28 Na+ are either in SIII’ or SIII. This marks a first singular difference between NaY and NaX zeolites regarding CO adsorption.

ACS Paragon Plus Environment

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 2 of 7

As FTIR spectroscopy is among the most molecularly informative techniques for gas/cation interaction, one might expect to distinguish CO/Na+ interaction in connection with a sitespecific criterion, provided accessible SII, SIII and SIII’ sites are concerned. In other words, singular FTIR CO signatures should reflect the type of S-sites that Na+ cations occupy. Besides this site-specific criterion, another criterion can be also highlighted. As discussed in the FTIR literature for NaY, the progressive CO adsorption up to the highest CO pressures results from the completion of the SII Na+ cations coordination vacancies.3-6 Accordingly, mono, di- and tri-carbonyls could be formed as carbonyl adsorbed species -Na+(CO), Na+(CO)2 and Na+(CO)3 have been evidenced through their respective vibrating average νCO positions ca 2171, 2163 and 2150 cm-1, respectively.3,5 As a result, IR allows us to also consider complex-specific CO as an additional criterion. Given the both site-specific and complex-specific criteria, NaY is the good candidate for focussing only on the complexspecific criterion because only SII sites are occupied. As a result, regarding the COcomplexes with pressure dependence, polycarbonyls (di- and/or tri-) are most likely for the highest CO pressures.3,5 Surprisingly, Nour et al. found that Na+ in SII only coordinate with up to ~ 0.9 CO, Table 4. In other words, monocarbonyls would be the more likely formed CO-complex even for the highest CO pressures. This fully contradicts the FTIR literature for NaY regarding CO-complexes with pressure dependence. Moreover, to justify that only monocarbonyls form throughout CO admittance, Nour et al. wrote ‘Such an adsorption behavior is consistent with previous experimental findings reported by Hadjjivanov et al. on the CaNaY Faujasite’ This assertion is somewhat confusing as Hadjjivanov et al. referenced work2 dealt with CO interacting with Ca2+ cations in Y Faujasite. Results obtained with Ca2+ do not seem to us as immediately transferable to the case of Na+. This quoted hardly proves that only one CO binds with one Na+ whatever the CO pressure and counter the conclusions of recent published surveys. 3,5 As for whether the experimental FTIR results made for the low temperature range [77-100] K are transferable to Nour et al.’s NaY survey as they claimed, it clearly seems that the CO-complexes that are formed for medium to higher CO pressures are more likely polycarbonyls. It also works well when passing to other cations like Ca2+ as Hadjjivanov et al. reported and this fully contradicts both the results and assertions Nour et al. made when citing previous FTIR results. Let us now discuss the two coordination modes i.e., C-end or O-end complexes with Na for monocarbonyl species. From the FTIR literature, it is well established for the low temperature range [77-100] K, adsorption of CO on cationic zeolites is mainly through a C-end formation. Nonetheless, Tsyganenko et al. previously evidenced using variable-temperature FTIR


2

Page 3 of 7

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


spectroscopy: these C-end or O-end two coordination modes are temperature dependent.6 As a result, the higher the temperature, the more O-end complexes will be produced. To illustrate, at room temperature (300K), Tsyganenko et al. calculated about 35% of the monocarbonyl species to be O-end complexes for NaY. As Nour et al.’s simulations were realized at 300K for the NaY-CO system it would have been of fruitful interest to discuss this point in connection with Tsyganenko et al.’s previous results. Unfortunately, Tsyganenko et al.’s elegant work was neither considered nor quoted whereas Nour et al. discard any O-end complexes for NaY at 300 K. So, let us discuss on it briefly. Tsyganenko et al. reported that monocarbonyls are temperature dependent regarding a C-end or an O-end binding. From the lowest temperature (i.e., 80 K) to higher temperatures, this feature results from an endothermic isomerisation process that engages the two modes of coordination and can be described by the following isomerisation equilibrium (1): (1)

ZNa+…CO ↔ ZNa+…OC,

where Z stands for the zeolite lattice.

Accordingly to equilibrium (1), Tsyganenko et al. found ∆H°(1) = + 2.4 kJ/mol for the enthalpy value of this isomerisation process. Let us now go back to Nour et al.’s results regarding a C-end or an O-end binding. They discarded the O-end binding complex at 300 K as they found this O-end complex would result in ‘an adsorption enthalpy of −19.0 kJ/mol much lower than the values reported above’ i. e. -23.0 kJ/mol for a C-end complex, a value they qualified as being ‘in very good agreement with the previous experimental and theoretical values that span in the range [−21.0, −25.0 kJ/mol]. In other words, the following equilibrium (2) seems to us to be more adequate with experimental enthalpy values: (2)

ZNa+ + CO ↔ ZNa+…CO

with ∆H°(2) = -23 kJ/mol

rather than an O-end binding hypothesis as described by the following (3) equilibrium: (3)

ZNa+ + CO ↔ ZNa+…OC

with ∆H°(3)

Then, from Nour et al.’s claim, equilibrium (3) must be discarded to account for CO adsorption on NaY at 300K as the ∆H°(3) value they calculated for a O-end complex (i.e. −19.0 kJ/mol) is too low compared with previous results spanning from −21.0 to −25.0 kJ/mol. Nevertheless, taking into account both Tsyganenko et al.’s result regarding the isomerisation process (1) and the above equilibrium reactions, it is straightforward to say that equilibrium (3) is the sum (1) + (2). As a result, we calculate a higher ∆H°(3) = -20.6 kJ/mol a


3


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 7

resulting value (-23 + 2.4) to be compared with the −19.0 kJ/mol Nour et al. obtained and finally rejected as an acceptable value. Comparison of the above-calculated ∆H°(3) = -20.6 kJ/mol based on Tsyganenko et al. findings with previous experimental values (i.e., −21.0, −25.0 kJ/mol) makes the O-end mode more energetically acceptable in respect to the precision Nour et al. agreed with for a C-end mode (4 kJ/mol range of precision). Therefore, Nour et al.’s results lack conclusions for a Cor an O-end mode of CO coordination to Na+ in the NaY at 300 K because they did not take into consideration Tsyganenko et al.’s previous results evidencing isomerisation process.

The second point we would like to discuss is the 20 T cluster Nour et al. chose as a prerequisite realistic model for further GCMC simulations. Whatever the SII or SIII’ positions for Na+, their 20 T cluster includes: ‘Only one Al atom was included[…] in the present models either in a six-membered or in a four-membered ring close to the Na+ at the SII site (Figure 1a) and the Na+ at site III′ (Figure 1b) respectively.’ Compared to a previous DFT study dedicated to the NaY/CO system,7 the Nour et al.’s 20 T model with only one Al atom would strongly orient the nature of the Na+/CO complex. Indeed, it was demonstrated that the aluminum content of 6-MR is a key parameter for the formation of either mono-, di- or tricarbonyl species.7 The following conclusions were drawn from a comparison of experimental FTIR results and the different AlmSi(6-m)NamO18H12 clusters (m = 1,2 or 3 ) and n-CO as Na+ligand (with n = 1,2,3) assessed.7 First, whatever the number of Al atoms (1, 2 3) for the 6MR cluster, monocarbonyls’ formation is well reproduced if Na+-CO bonding length, adsorption energy and matching comparison with the IR experimental νCO stretching vibration are taken as criteria. However, when the type of Na+/CO complex is focused to account for the CO loading (i.e., when an increase of CO pressure is considered), the one Al cluster failed to match experimental observations and findings (i.e., formation of polycarbonyls).7 To match well with the formation of di- or tri-carbonyl species, clusters with two or three Al atoms are needed. As a consequence, the one Al atom model of Nour et al. is too selective and too constraining because monocarbonyls would be the only CO complexes to be formed for NaY.7 The second aspect to be discussed regards the Si/Al ratio dependence for the DFT study.7 As outlined above, monocarbonyls’ formation would not be influenced by the Si/Al ratio criterion instead of polycarbonyls. If we consider the Si/Al ratios of NaY and NaX (i.e. 2.4 and 1.1, respectively) in connection with cluster models describing the 6-MR and 4-MR (i.e., SII and SIII’) the one Al atom cluster Nour et al. chose, whatever the S-site, is a questionable


4

Page 5 of 7

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


choice. While acknowledging that for the NaX (Si/Al = 1.1) the one Al cluster model is a coherent choice for either SII or SIII’ site (according to both the Lowenstein’s Al–O–Al avoidance rule14 and the presence of a Na+ cation thus requiring at least one Al atom), this is not by far the most appropriate model to assess the difference of Si/Al ratios when passing from NaX to NaY. Indeed, for the NaY with a Si/Al ratio close to 2.6, most of the Si is surrounded by two or one Al atoms (c.a 40% and 37%, respectively).15 Comparison with IR spectra from CO adsorption on NaY (77 K) has led Huber and Knözinger to the corroborative conclusion that 6-MR having 2 Al are the most abundant. 16 Comparison with the 1.1 and 2.4 Si/Al ratio of NaX and NaY, respectively, the one Al atom cluster Nour et al. chose seems us a lot questionable as it does not match with previous results15-16 and would hardly account for any expected differences for the two X or Y zeolites. Moreover, further CGCM simulation results lead Nour et al. ‘to the conclusion that the Na+ SIII′ can coordinate 2 CO molecules in the intermediate and high pressure domains (Figures 4b,c)’. If they had considered a cluster with more than one Al atom for SII, their conclusion (i.e., only monocarbonyls as adsorbed species for NaY) would be different. Unfortunately, neither more Al-rich clusters nor previous studies dealing with this were considered in their work.

The third point to discuss involves the quantitative results from CGCM simulations and the conclusions regarding CO loading in NaY that Nour et al. made. They wrote ‘that the Na+ SII can coordinate with up to ~0.9 CO molecule in its first coordination sphere for a loading of 49 CO molecules/u.c. corresponding roughly to the saturation capacity. Such a result also clearly shows that there is a significant fraction of CO molecules that do not interact directly with the Na+SII.’ We acknowledge that CO loading in NaY should distinguish adsorbed species that are’ truly Na+-coordinated’ from those that are adsorbed in a condensed state in the confined micropores as we previously reported.8 This seems also to work with other weak-binding gases like dinitrogen (N2) as well as CO/N2 mixture when it loads NaY.9,10 However, the above-mentioned quantitative results Nour et al. report, should be looked at carefully. First, because Nour et al. have considered monocarbonyls as the sole adsorbed species through Na+-coordination, a result that seems unlikely as discussed above. Second, when examining the absolute adsorption isotherm for CO they reported for NaY in Figure 6. Indeed, from the CO isotherm in Figure 6,1 it appears that the saturation can hardly be claimed as roughly attained because neither a plateau nor one at its start-shape is clearly observed for the pressure conditions they explored. This feature contrasts with the more


5


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 7

reliable conditions that previous studies dealing with CO or N2 isotherms have chosen for deducing quantitative results.8,9 Consequently, Nour et al.’s quantitative results drawn for the NaY (at least) should be reconsidered.

The fourth point we will discuss regards many incorrect assertions and comparisons Nour et al. made when referencing the literature. First, regarding geminal species (i.e., dicarbonyls) that could be formed in the NaX, they wrote that their formation ‘has been already identified by experimental measurements realized in the low temperature range [80−100] K in this NaX system.52,69−72’ As far as we are aware, only Martra et al.’s work referred to NaX,4 (quoted as Ref. 69 in Nour’s paper) whereas others (quoted as Ref. 69 to 72) referred to NaY. Interestingly, Martra et al. claimed that according to the intensity of the IR bands they observed for NaX/CO this feature ‘indicates that SIII positions are less populated than SII.’4 Surprisingly, this deduced-insight fully contradicts the argument Nour et al. developed all along their paper regarding SII and SIII’ occupancy. Let us take the example of another misunderstood argument. To justify ‘the formation of a few species having a peculiar arrangement of CO, where the adsorbate molecules interact with both Na+ SIII′ via its C atom and a neighbor Na+ SII by means of its O atom (Figure 4d)’, Nour et al. referred to other previous DFT studies including ‘NaY27 and CuNaY29 and other types of alkali- and alkali earth-zeolite such as Na- and K-ZSM5,38 Na-,37 K-73 and Mg74

FER, and Na-LTA.55’ In view of both the different structure types of zeolites and the

different cations they referred to it would mean the ‘double interaction’ between SII and SIII’ would not be so specific for NaY and NaX systems regarding CO adsorption. In other words, this ‘double interaction’ would not require the SII and SIII’ specific arrangement in Faujasite. We are somewhat doubtful of such a comparison. In fact, going on ahead in their text, this double interaction in NaY appears as a prerequisite result for easily introducing ‘migrations of a fraction of Na+ SII to vacant SIII′ sites.’ Particularly to strengthen justification of the adsorption enthalpy profile they obtained for CO loading with coverage dependence as they calculated for NaY, (Figure 6, right part). Surprisingly, such a double interaction involving both SII and SIII’ as well as CO via its O- or C-end could resemble a ‘pseudo’ dual-site complex. Note however that ‘real’ dual site complex formation has been found unlikely for NaY as recently discussed and assessed.11-13 If ‘pseudo’ dual complex were to be an à priori hypothesis (although the ‘real’ dual complex was discarded) it would require stronger arguments than those Nour et al. developed within their paper.


6

Page 7 of 7

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


Additionally, many of Nour et al.’s other conclusions are also rather inconsistent with the NaX literature. All in all, regarding published literature for Na+ and other cationic X or Y zeolites systems interacting with CO, additional researches are required on both the experimental IR side and the theoretical approaches.

To conclude, we hope that the different points we have outlined will help advance both the reader’s understanding and general discussion of the topic towards solving the complex issue of NaY and NaX interactions with CO. Experimental literature, including other previous surveys combining both experimental and modeling aspects, must be reported with precision as a prerequisite for useful theoretical investigations.

References:

(1) Nour, Z.; Berthomieu, D.; Yang; Q., Maurin, G. J. Phys. Chem. C 2012, 116, 24512. (2) Hadjiivanov, K.; Knözinger, H.; Ivanova, E.; Dimitrov, L. Phys. Chem. Chem. Phys. 2001, 3, 2531. (3) Cairon, O.; Loustaunau, A. J. Phys. Chem. C 2008, 112, 18493. (4) Martra, G.; Ocule, R.; Marchese, L.; Centi, C.; Coluccia, S. Catalysis Today 2002, 73, 83. (5) Cairon, O. Phys. Chem. Chem. Phys. 2010, 12, 14217. (6) Tsyganenko, A. A.; Platero, E. E.; Areán, C.O.; Garonne, E.; Zecchina, A. Catalysis Letters 1999, 61, 187. (7) Cairon, O.; Guesmi, H. Phys. Chem. Chem. Phys. 2011, 13, 11430. (8) Cairon, O.; Bellat, J. P. J. Phys. Chem. C 2012, 116, 11195. (9) Cairon, O. Phys. Chem. Chem. Phys. 2012, 14, 12083. (10) a) Cairon, O. J. Phys. Chem. C 2012, 116, 25949; b) Cairon, O. ChemPhysChem. 2013, 14, 2744. (11) Nachtigall, P.; Delgado, M.R.; Nachtigallova, D.; Arean, C.O. Phys. Chem. Chem. Phys. 2012, 14, 10353. (12) Cairon, O. Phys. Chem. Chem. Phys. 2012, 14, 10351. (13) Nachtigall, P.; Delgado, M. R.; Nachtigallova, D.; Arean, C.O.Phys. Chem. Chem. Phys., 2012, 14, 1552. (14) Loewenstein, W. Am. Mineral. 1954, 39, 92. (15) Melchior, M.T., Vaughan, D.E.W., A.J. Jacobson, J. Am. Chem. Soc. 1982, 104, 4859. (16) Huber, S., Knözinger, H. Applied Catalysis A: General 1999, 181, 239.


7

A Computational Exploration of the CO Adsorption in Cation

Recommend Documents