Ammonia pools in zeolites for direct fabrication of catalytic centers

Reduction process is a key step to fabricate metal-zeolite catalysts in catalytic synthesis. However, because of the strong interaction force, metal oxides in zeolites are very difficult to be reduced. Existing reduction technologies are always energy-intensive, and inevitably cause the agglomeration of metallic particles in metal-zeolite catalysts or destroy zeolite structure in severe cases. Herein, we disclose that zeolites after ion exchange of ammonium have an interesting and unexpected self-reducing feature. It can accurately control the reduction of metal-zeolite catalysts, via in situ ammonia production from ‘ammonia pools’, meanwhile, restrains the growth of the size of metals. Such new and reliable ammonia pool effect is not influenced by topological structures of zeolites, and works well on reducible metals. The ammonia pool effect is ultimately attributed to an atmosphere-confined self-regulation mechanism. This methodology will significantly promote the fabrication for metal-zeolite catalysts, and further facilitate design and development of low-cost and high-activity catalysts.

or hexachloroplatinic acid, the zeolite needs to be pretreated by calcination in oxygen or air before reduction. Large agglomeration of Pt particles occurs at external surfaces of the zeolite particles, instead of inside micropores. I am not saying that the APE is not good for Cu, Pd and Ag, but I think that the effect is difficult to generalize to other systems and experimental conditions. There are other reducing agents better than ammonia, depending on metals. For example, alcohol is a good reducing agent to support gold nanoparticles on zeolite. Even water vapor coming from dehydration of zeolite can a reducing agent in the case of gold. Overall, I do not recommend this paper for publication in NatureComm. This paper looks interesting and suitable for a catalysis journal, but it seems lacking general attention.
According to your suggestion, we supplemented more valuable characterizations 23 and found some new merits of our APE method. The added characterizations involved 24 in-situ TEM, in-situ XRD, TG-DTA-MS, SEM-EDS elemental mapping, XPS, H2-TPR, 25 and TG. In addition, besides the as-proved metals like Cu, Ag, and Pd, more base metals 26 (Fe, Co, Ni) and noble metals (Pt and Au) are also proved their easier reduction by our 27 APE.

28
• We reconducted the TEM characterization of Cu-MOR and supplemented new 29 TEM images of Ag-ZSM5, Pd-β, and Pt-MOR, the corresponding results are 30 shown in the followed Fig. 1, Fig. 2, Fig. 3, and Fig. 4, respectively, as well as 31 in Fig. S9, Fig. S19, Fig. S20, and Fig. S21, respectively, in the new SI of the 32 revised manuscript. Based on the results given in Figs. 1-4, it is obvious that 33 the metal particles inside zeolite fabricated by APE are much smaller than 34 those by traditional H2 reduction. In addition, the metal particles after H2 35 reduction are easier to migrate to the outer surface of zeolite, but the metal 36 particles fabricated via APE are confined in the zeolite. This should be one of 37 the reasons why the size of metal particles generated by traditional H2 38 reduction is larger than that by APE.  elements mapping is weaker than that of metal-zeolite (H2) samples, 55 illustrating that there are more metal particles on the surface of metal-zeolite 56 (H2) samples than that of metal-zeolite (APE) samples. This finding is 57 consistent with the result indicated by TEM in the above Figs. 1-4. Therefore, 58 we can conclude that our APE can effectively confine the reduction and 59 formation of smaller metals nano particles inside the zeolite, better than the 60 traditional H2 reduction.   • Thermogravimetry-differential thermal analysis and mass spectrometry (TG-80 DTA-MS) characterization was supplemented to in-situ trace the formation of 81 gas-phase product during APE reduction process. The differential thermal 82 variation and gas-phase product were collected and analyzed in real time by 83 DTA and MS, respectively. The result is displayed in the followed Fig. 9, as 84 well as in Fig. S4   was heated to 500 ℃, and the XRD pattern was recorded in the temperature 97 range of 300-500 ℃. Copper characteristic peak was not observed, because 98 of the high dispersity of copper particles. The framework of MOR zeolite was 99 also stable, certifying that both the high-temperature calcination and the 100 copper particles formation could not destroy the crystalline phase of MOR 101 zeolite. peaks, and no metal diffraction peak is observed in all samples, implying that 116 the MOR crystal phase structure of all samples is integrated, and the degree 117 of metal dispersion is high. According to Figs. 12 and 13, results both prove the superior reduction ability and wide applicability of APE, 119 not only being suitable for base metals but also for noble metals. In all samples, 120 more than 50% reduction degree of metal species can be obtained by APE 121 method. It's worth mentioning here that zeolite dehydration occurs in the 122 temperature-rise period of calcination. It is a fact that the water vapor coming 123 from dehydration of zeolite can reduce Au. However, as proved by followed 124 Fig. 12e, APE realizes a much higher reduction degree (92% vs 31%) than the 125 water vapor reduction in the case of Au. Intensity (a   5  10  15  20  25  30  35  40  45  50  55  60  65

2(degrees)
Intensity ( Reduction degree by zeolite dehydration = 31% = d, Pt and e, Au) samples with or without APE reduction. Note: Reduction degree = 1-(the H2 130 consumption of APE sample / the H2 consumption of the sample without reduction) (1 -(area of blue curve / area of 131 red curve)). The Au(1.0wt%)-MOR(Ar) sample refers to the sample reduced by zeolite dehydration.  coke in the methane coupling reaction occurring at 800 ℃ (see Fig. 15). These 147 findings suggest that the superior anti-carbon ability of APE, not only works 148 well on medium-temperature reaction (DME carbonylation, 220 ℃) but also 149 takes effect on high-temperature reaction (methane coupling, 800 ℃).  degree (see Table S1).' In our work, we had employed five different zeolites and three metals to verify the 234 general applicability of APE. Three unsimilar reactions, whose reaction temperatures 12d and e). In control experiment, Au(1.1wt%)-MOR(Ar) was synthesized via 253 calcination in Ar atmosphere, and in-situ reduced by zeolite dehydration. It is a fact that 254 the water vapor coming from dehydration of zeolite can reduce Au. However, as proved 255 by followed Fig. 12e, APE realizes a much higher reduction degree (92% vs 31%) than 256 the water vapor reduction in the case of Au.

257
The XPS results revealed that, after APE reduction, all the Fe(1.8wt%)-  This indicates that the noble metals (Pt and Au) on the MOR were easier to be reduced 264 than the transitional metals (Fe, Co, and Ni).

265
H2-TPR and XPS results both prove the superior reduction ability and wide 266 applicability of APE, not only being suitable for base metals but also for noble metals.  Au) but also on base metals (Fe, Co, Ni).

294
In addition, we also further proved the superiority of APE in encapsulating of 295 metal particles inside zeolite. We reconducted the TEM characterization of Cu-MOR 296 and supplemented new TEM images of Ag-ZSM5, Pd-β, and Pt-MOR, the 297 corresponding results are shown in the followed Fig. 1, Fig. 2, Fig. 3, and Fig. 4, 298 respectively, as well as in Fig. S9, Fig. S19, Fig. S20, and Fig. S21 confined in the zeolite. This should be one of the reasons why the size of metal particles 304 generated by traditional H2 reduction is larger than that by APE. agglomeration of copper particles, but APE reduction is a successful way to inhibit it. Hence, it is a fact that we discovered and contributed a brand-new APE, by which 319 to reduce metal nanoparticles inside zeolite channels much better, more quickly and 320 more conveniently, overcoming a lot of existing problems faced by us, such as 321 calcination, high temperature, energy-consuming, sintering, etc.  Fig. 14 and Fig. 15 here, as well as in Fig. S24 and catalyst prepared by the traditional H2 reduction formed a lot of hard coke in the 334 methane coupling reaction occurring at 800 ℃ (see Fig. 15). These findings suggest 335 that the superior anti-carbon ability of APE, not only works well on medium-336 temperature reaction (DME carbonylation, 220 ℃) but also takes effect on high-337 temperature reaction (methane coupling, 800 ℃). The TEM analysis and copper particle size distribution of fresh Cu(3.41wt%)-  zeolite. In addition, the size of Pt nanoparticles obtained by APE is also smaller than 368 that by traditional H2 reduction. Therefore, the results confirmed the general 369 applicability of our APE again. Moreover, the APE is also easy to generalize to other 370 metals. As in followed Figs. 2 and 3, it is obvious that the Ag and Pd particles inside 371 zeolite fabricated by APE are also much smaller than those by traditional H2 reduction. Ni(1.2wt%)-MOR(APE) reached 86%, 51%, and 67%, respectively (Fig. 12a-c).

383
The XPS results revealed that, after APE reduction, all the Fe(1.8wt%)-  (Fig. 13a-e). It clearly proved that these metal samples were reduced 387 by the APE process.

388
In summary, all results prove the superior reduction ability and wide applicability 389 of APE. example, alcohol is a good reducing agent to support gold nanoparticles on zeolite.

393
Even water vapor coming from dehydration of zeolite can a reducing agent in the case 394 of gold.

395
Thank you very much for your comments. In this paper, we presented a new and 396 efficient APE for metal-zeolite catalysts preparation, which is not in conflict with the 397 known ways those you suggested.

398
Here, firstly, we want to emphasize that our work does not use ammonia to reduce 399 metal but find that NH4-zeolite has potential self-reduction ability to the encapsulated 400 metals. Secondly, it is true that polyvinyl alcohol (PVA) can reduce Au nanoparticles in 401 some cases, as declared by this reviewer, but it is difficult to reduce more metals. In 402 addition, the employed expensive PVA and the co-produced wastewater by using this 403 method are potential problems too. In contrast, the APE in our work is greener, simpler, 404 more economical, and environmentally friendly. It provides one more choice for metal-405 zeolite catalyst design, preparation, and application.  Reduction degree by zeolite dehydration = 31% = d, Pt and e, Au) samples with or without APE reduction. Note: Reduction degree = 1-(the H2 419 consumption of APE sample / the H2 consumption of the sample without reduction) (1 -(area of blue curve / area of 420 red curve)). The Au(1.0wt%)-MOR(Ar) sample refers to the sample reduced by zeolite dehydration.

<b>REVIEWERS' COMMENTS</b>
Reviewer #1 (Remarks to the Author): This paper is much better than the original. The core idea, that ammonium ions in zeolites form ammonia that gently reduces metal cations in zeolites is satisfactorily demonstrated by the data. The presentation is more prudent than before, but it would still be a very good idea for the authors to pull back on statements such as that CO and H2 are "explosive and poisonous." There is no insight there, and in practice safe work is done with these gases. Referring to catalysts as "more powerful" loose, overwrought language and should be removed. Even the term "ammonia pools" is inappropriate and should be avoided. The evidence of the intermediate metal oxidation states is not very strong--other techniques, such as spectroscopic techniques, would provide a stronger foundation. The catalyst preparation data are the interesting new results. The catalyst performance results are superficial and do not link strongly with the preparation data and would be best removed, leading to a shorter, sharper paper. Overall, this work contains something new. It does not match up with the strongest papers in the journal, but it is as good as the weaker ones.
Reviewer #3 (Remarks to the Author): In this paper titled by "Ammonia Pools in Zeolites for Direct Fabrication of Catalytic Centers", an original concept of 'Ammonia Pools in zeolites' was proposed. The named 'ammonia pool effect' (APE) is very interesting. Zeolites play important roles in the areas of chemistry and environment, but preparing metal-zeolite catalysts without post-treatments (such as erasing reduction by hydrogen under higher temperature), in a simpler and more environmentally friendly way, is still a challenge both in science and engineering. I believe that the APE contributed by this paper solved this problem in metal-zeolite catalysts preparation.
I checked all the materials submitted by authors, including the supporting information. This paper is well organized. The high-quality TEM images support the newly discovered encapsulation and confinement effects of APE to metallic nanoparticles. In comparison with traditional H2 reduction, the APE in this paper is more powerful, feasible and scalable, being available to almost all zeolites and metals.
The authors also verified that the APE could encapsulate metallic Pt into zeolites successfully, where the traditional methods cannot realize in fact. Especially for the gold reduction, the gold reduction degree obtained by APE was 92%, much higher than the 31% by previously reported zeolite dehydration. In my opinion, the excellent in-situ reduction ability and broad application of APE for zeolites and noble/base metals are very promising, even in future plant engineering and commercialization. Therefore, I recommend this paper to be accepted for publication by Nature Communications after minor revision.

Few comments are listed below for authors:
In the section of Structural characterization, the recording method of in-situ TEM images is not clear. The patterns were recorded with the increase of temperature or at a fixed temperature?
A feed gas containing hydrogen was used to investigate the catalyst stability in carbonylation reaction. What is the purpose of hydrogen in feed gas? We are very grateful to you for assessing our revised manuscript. We acknowledge the reviewers for your time and expertise. We are very appreciated to the reviewers for your constructive comments and suggestions. You help us to significantly improve our manuscript. Point-by-point responses to the remaining concerns of your comments are displayed as follows:

Journal
Q1. The presentation is more prudent than before, but it would still be a very good idea for the authors to pull back on statements such as that CO and H2 are "explosive and poisonous." There is no insight there, and in practice safe work is done with these gases.
Referring to catalysts as "more powerful" loose, overwrought language and should be removed. Even the term "ammonia pools" is inappropriate and should be avoided.
We are very thankful to you for your professional comment. According to your suggestion, the statements such as that CO and H2 are 'explosive and poisonous' have been thoroughly pulled back in the revised manuscript. The loose, overwrought language such as 'more powerful' has been removed completely. However, for the term of 'ammonia pools', we think it will be better if it can be reserved, because this concept is the core of our work and can image the phenomena. This term is also considered to be an original concept in zeolites, and other reviewers also think that the ammonia pool effect (APE) is very interesting and suitable.

Q2. The evidence of the intermediate metal oxidation states is not very strong, other
techniques, such as spectroscopic techniques, would provide a stronger foundation.
We are very thankful to your professional suggestion. According to your comment, we supplemented another spectroscopic technique, Cu LMM auger electron spectroscopy (AES), to provide a stronger foundation and evidence of the intermediate metal oxidation states. The added characterizations are displayed in the Supplementary Fig.   5b and Fig. 5c and are also shown here as Fig. 1a and Fig. 1b as below.   Fig. 2d and Table S1. The lower Cu 0 content was not in favor of DME carbonylation. In addition, the monovalent copper species on the Cu + (3.71wt%)-MOR also inhibited DME carbonylation (Table S4).H-Cu(X)-MOR(APE) was lower than that of H-Cu(X)-MOR(H2) (Fig. 2d and Supplementary Table 1). These phenomena allow us to consider that the synergy of Cu 0 and Cu + may be the key factor in significantly promoting the DME carbonylation. Our characterization results also revealed that H-Cu(X)-MOR(APE) samples formed high ratio of Cu + to Cu 0 (Table S1). Moreover, In addition, the H-Cu(3.41wt%)-MOR(APE), with the highest ratio of Cu + (40.6%) to Cu 0 (45.8%) (See Supplementary Table 1), generated the best performance of DME carbonylation. Thus Hence, we confirm that the coexistence and synergy of Cu 0 and Cu + played the key role in enhancing the DME carbonylation of H-Cu(X)-MOR(APE).' Overall, this work contains something new. It does not match up with the strongest papers in the journal, but it is as good as the weaker ones.
We are very grateful to you for assessing our revised manuscript. We acknowledge the reviewers for your time and expertise. We very much appreciate the reviewers for your constructive comments and suggestions. You help us to significantly improve our manuscript. We are very grateful to you for assessing our revised manuscript. We acknowledge the reviewers for your time and expertise. We very much appreciate the reviewers for your constructive comments and suggestions. You help us to significantly improve our manuscript. Point-by-point responses to remaining concerns of your comments are displayed as follows:

In this paper titled by 'Ammonia Pools in Zeolites for Direct Fabrication of Catalytic
Few comments are listed below for authors:

Q1. In the section of Structural characterization, the recording method of in-situ TEM
images is not clear. The patterns were recorded with the increase of temperature or at a fixed temperature?
We are very thankful to you for the professional suggestion. According to your suggestion, we have supplemented the details of the recording method of in-situ TEM images, the revised text is also displayed as follows: 'In-situ TEM was conducted by JEM-2100F: the temperature of H-Cu(3.27wt%)-MOR(H2) sample was heated to 500 ℃ in the air with a rate of 10 ℃/min. The TEM images were recorded at the fixed temperatures of 20 ℃, 100 ℃, 200 ℃, 300 ℃, 400 ℃, and 500 ℃, respectively. When the temperature rose to 500 ℃, it was kept for 40 min, and the images were recorded every 10 minutes.' As shown, the patterns of the in-situ TEM were recorded at a fixed temperature.

Q2.
A feed gas containing hydrogen was used to investigate the catalyst stability in carbonylation reaction. What is the purpose of hydrogen in feed gas?
We are very thankful to you for your professional question. Although the modification of pyridine is an effective measure to prepare a long-life MOR zeolite catalyst in DME carbonylation. However, with the increase of reaction duration, Py-MOR catalyst still behaves inevitable deactivation due to the carbon deposition derived from the side reaction of MTO in 12-MR of MOR zeolite. The addition of hydrogen in feed gas can effectively inhibit the deactivation of zeolite catalysts in the long-time running process, because the formation of coking precursor from the side reaction of MTO can be interdicted when there is hydrogen. We added hydrogen in feed gas to gain a long and stable catalytic performance of Py-Cu(3.41wt%)-MOR(APE) catalyst.