Thermally Tunable Glass Foams with Controllable Pore Size via

10 hours ago - Glass foams are being widely used as constructional materials due to their unique properties in thermal insulation, fire retardation, a...
0 downloads 0 Views 3MB Size
Subscriber access provided by Kaohsiung Medical University

Article

Thermally Tunable Glass Foams with Controllable Pore Size via Network Manipulation: A Novel Meltcasting and Float-manufacturable Glass Foaming Method Hongwei Guo, Congcong Zhao, Yuxuan Gong, Jian Xu, Mariano Velez, and Relva C. Buchanan ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.8b01225 • Publication Date (Web): 06 Jun 2018 Downloaded from http://pubs.acs.org on June 6, 2018

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 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 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.

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 36 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 Sustainable Chemistry & Engineering

Thermally Tunable Glass Foams with Controllable Pore Size via Network Manipulation: A Novel Melt-casting and Floatmanufacturable Glass Foaming Method Hongwei Guoa, Congcong Zhaoa, Jian Xub*, Yuxuan Gongc, d*, Mariano Velezc, and Relva C. Buchanand a

School of Materials Science and Engineering, Shaanxi University of Science and Technology, 111 County Rd, Xi’an 710021, China

b

School of Physics and Electronic Information, Henan Polytechnic University, 1 Heping St, Jiaozuo 454000, China

c

Department of Mechanical and Materials Engineering, College of Engineering and Applied Science, University of Cincinnati, 2600 Clifton Ave, Cincinnati, Ohio, 45221, United States

d

Glass Coatings and Concepts LLC, a subsidiary of the Shepherd Materials Science Companies, 300 Lawton Ave, Monroe, Ohio, 45050, United States

Keywords: glass foams; network manipulation; thermal properties; pore size; float-manufacturable *Corresponding author: Dr. Yuxuan Gong and Dr. Jian Xu Tel.: +1-513-539-5300 (Dr. Yuxuan Gong); +86-391-3987811 (Dr. Jian Xu) Fax: +1-513-539-5177 (Dr. Yuxuan Gong); [email protected] (Dr. Jian Xu) Email: [email protected] (Dr. Yuxuan Gong)

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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

Abstract Glass foams are being widely used as constructional materials due to their unique properties in thermal insulation, fire retardation, and shockwave absorption. However, the cost of energy consumption and processes in a conventional glass foam production limited the use of glass foams as sustainable materials. In this study, for the very first time, thermally tunable CaO-SnO2-P2O5-SiO2 glass foams with controllable pore size were presented as a novel category of melt-casting and float-manufacturable glasses. It was found that the pore size and thermal properties become tunable by manipulating the glass network, i.e. connecting linear chained Sn-P network with [SiO4] units. In addition, the unique combination of thermal properties and porous structure of CaO-SnO2-P2O5-SiO2 glasses shows potential in float glass foam production, which can produce glass foams sheet-by-sheet with less complexity in manufacturing processes. Introduction Glass foam is a category of lightweight constructional materials that has been widely used in the exterior and interior wall of buildings due to its thermal and acoustic insulating properties 1-3. The porous structure, either being closed or open pores, of glass foam provides its unique properties in thermal insulation, fire retardation, and shockwave absorption

2, 4-6

. However, the unit price of glass foam limited its use in

constructional needs, and tends to further restrict its use when the needs for energy-efficient buildings becomes more demanding 6. The production cost of glass foams is associated with the investment of production equipment, large energy consumption, and sometimes cost of raw materials 2, 6-7. During a typical production process, glass foams are being made by batching and melting of glasses with desired stoichiometry (batching and melting are not necessary when waste glasses are being used), grinding the as-melted glass or waste glasses (e.g. cathode ray tube, window or bottle glass), mixing the glass particulates with foaming agents, foaming the mixture (in a mold) at elevated temperature, and annealing the as-prepared glass foam

ACS Paragon Plus Environment

Page 2 of 36

Page 3 of 36 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 Sustainable Chemistry & Engineering

through a lehr 5. Finally, glass foams are cut or machined into desired size prior to being used in construction sites. Couple of approaches has been applied to lower the cost of glass foam production, among those include the use of waste glass cullets

5, 7

, novel foaming agent from municipal wastes (e.g.

slag, fly ash, porcine bone, egg shell, and etc.) 2-3, 8-9, and fast sintering schedule 2. The use of waste glass cullets somehow allows a melting-free processing of producing glass foam, but also introduces accumulation of undesirable element (e.g. iron, and sulfides) and fluctuations in batch-to-batch composition (i.e. inconsistency in batch-to-batch composition of glass cullets) 6, 10-11. Similarly, the use of novel foaming agents from wastes only helps with the cost of raw materials. Methods of producing glass foams at lower energy consumption and higher production efficiency (or a much more automated production) have yet to be reported. In addition, the foaming process and the final product were reportedly to be sensitive over glass composition as well as the chemistry of foaming agents

5, 10-11

. For example, a decomposition type

foaming agent (e.g. carbonates, such as Na2CO3) usually forms open porosity in the glass foam; also, the ion exchange between sodium in the foaming agent and alkalis in the glass network has been reported to change the viscosity, surface tension, and final pore structure of glass foams

5, 11

. Similarly, a redox-type

foaming agent like SiC will change the glass chemistry by introducing extra Si and reducing metallic cations in glass network (e.g. Zn, Pb, or Cu)

4-5

. Particularly, a study on the foaming agents shows that

superior properties of glass foams can be achieved with the combination of MnxOy and carbon as foaming agents

12

. Therefore, the interaction between foaming agents and matrix glasses will greatly affect the

properties of glass foams such as morphology of porosity, distribution of porosity, thermal and mechanical properties, and more importantly create problems in product consistency. With all these being said, is there a one-step and mold-free process of producing glass foam with controlled stoichiometry? Can the glass foam being produced just the way as float glasses, in other words a sheet-by-sheet production? Can such a process lower the cost of energy consumption, reduce the cost of labor, and increase the efficiency and consistency of production?

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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

In this paper, we report, for the very first time, a melt-casting method of obtaining glass foam that satisfies the requirements for a float production process of glasses. Glass foams, at least in the Si-P glass family per our study, show potential in being produced sheet-by-sheet in a state-of-art one step process (i.e. float glass foam manufacturing). Also, the as-prepared glass foams showed controlled chemistry, distribution of porosity, pore size, and thermal property with good mechanical strength (> 10 MPa). Moreover, the glass foams were found to be tunable in their thermal properties and pore size via the manipulation of glass network. It was found that the measured properties show a linear regression as a function of Si4+/Ca2+. More importantly, the cost-efficiency and potential impact of the shown method are discussed in detail. Materials and methods Processing Table 1. Batched composition of glasses with varying amounts of silica (mol %) Composition SiO2 P2O5 SnO2 CaO (mol%) Sn1 5.5 47 10 37.5 Sn2 6.5 47 10 36.5 Sn3 7.5 47 10 35.5 Sn4 8.5 47 10 34.5 Chemically pure ammonium diphosphate (Fisher Scientific, USA), tin (IV) oxide (Fisher Scientific, USA), calcium carbonate (Fisher Scientific, USA), silica (Fisher Scientific, USA) powders were weighed (see Table 1), mixed in a rotating roller for 12 hours, placed into an alumina crucible and melted at 1100 oC for 1 hour. The composition of glasses was designed to vary the ratio of Si/Ca while not changing the amount of P2O5 and SnO2 (Sn/P less than 0.25), assuming that the pore size and thermal properties of foam glass changes as a function of Si-network connectivity (Si4+/Ca2+). The glass forming liquid was poured into a pre-heated (500 oC) and rectangular-shaped (~2 cm X 1 cm X 1 cm) graphite mold, annealed in a resistively heated oven for 1 hour, and subsequently cooled overnight. The as-melted glasses show porous structure with entrapped gases from the decomposition of ammonium diphosphate

ACS Paragon Plus Environment

Page 4 of 36

Page 5 of 36 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 Sustainable Chemistry & Engineering

and calcium carbonate. Therefore, the foaming process was achieved during the melting processes without the use of foaming agent or a further sintering process. The edges of the as-prepared glass foams were polished prior to the measurement of CTE. Triplicates were performed on the CTE measurement in order to generate standard deviations.

Phase identification and analysis of bulk composition X-ray powder diffraction was performed on all pulverized glass samples with a Rigaku Dmax X-ray diffractometer. A Cu kα source, an emission current of 10 mA, and a voltage of 30 kV were used in all measurements. Data was collected from 10o to 70o 2θ with a step size of 0.01o. Bulk compositions as well as the chemical environments of the as-prepared glass foams were measured using a PHI Versaprobe Xray photoelectron spectrometer. In order to obtain the bulk composition, the fracture surfaces (as fractured in the introduction chamber, pressure is ~ 10-6 torr) were analyzed. The quantitative analyses were performed using high resolution scans of Si2p, P2p, Sn3d5/2, and Ca2p peaks. Relative sensitivity factors of all elements (RSFSi2p=1, RSFP2p=0.60, RSFSn3d5/2=4.32, RSFCa=9.32 and RSFO=4.77) were derived from high resolution XPS scans of a vacuum-fractured Sn-P-Ca-Si oxide glass, assuming that the composition of glass fracture surface was identical to the bulk glass composition (obtained via ICP-OES and lithium metaborate digestion). Qualitative and quantitative analyses were performed with PHI MultiPak V8.0 software; peak positions of all scans were shifted/calibrated using adventitious alkyl C1s peak located at 284.60 eV (the standard deviation of peak positions is less than 0.1 eV). The peak area value of each high resolution scan XPS peak was fitted by the Shirley-model after automated background subtraction. Details regarding XPS analyses on glasses are being reported elsewhere

13

. The obtained elemental

composition was then converted into oxide composition, as listed in Table 2. The measured standard deviation of Si is 0.2 mol %, P is 0.1 mol %, Sn is 0.4 mol %, Ca is 0.2 mol%, and O is 0.9 mol%.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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 36

Table 2. X-ray photoelectron spectrometer measured composition of as-prepared glass foams (mol %) Composition (mol%) Sn1 Sn2 Sn3 Sn4

SiO2

P2O5

SnO2

CaO

Si4+/Ca2+

5.7 6.5 7.4 8.3

47.2 46.7 46.3 45.9

9.4 9.3 9.2 9.1

37.7 37.5 37.1 36.7

0.15 0.17 0.2 0.23

Thermal analysis, characterization of pores, and mechanical strength test The DSC analyses were carried out using a NETZSCH Jupiter STA 409 thermal analyzer with a heating rate of 15 K/min. The CTE (coefficient of thermal expansion) values were measured from the casted glass foams using an Orton 1412 dilatometer (equipped with quartz sample tube and push rod), and a heating rate of 10 K/min was applied to all measurements with a high purity alumina as the standard. The microstructural image of the fractured glass foams was captured using a Zygo® NewViewTM optical profiler (a white laser interferometer), and the analysis of images was performed using ImageJ. For statistical significance, ten images from each sample were used for the microstructural analysis (pore size and rather than the requirements of ASTM standard 2. In addition, the flexural strength (four point bending test) was measured at a cross-head rate speed of 2 mm/min using an 8874 axial-torsion fatigue testing system (Instron, USA). The samples used for four point bending test were prepared using the method discussed elsewhere 2.

ACS Paragon Plus Environment

Page 7 of 36 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 Sustainable Chemistry & Engineering

Results and discussion Phase identification

Figure 1 X-ray powder diffraction pattern of all foam glass samples. The missing of crystallization peak indicated the samples to be X-ray amorphous.

Figure 1 shows the powder diffraction of all measured foam glass samples. All measured patterns did not show any diffraction peaks, which often can be found due to the crystallization of bulk glass as well as partial decomposition of foaming agent. Therefore, the as-prepared glass foams were X-ray amorphous. It has been reported that a crystalline phase containing glass foam (so called glass-ceramic foam) shows better mechanical performance comparing to its non-crystalline counterpart

2, 4-5

. The crystalline phase

contained in a glass-ceramic foam hinders the propagation of cracks, and therefore reinforces the glassceramic foam

14

. However, uncontrolled crystallization of glass foams can be challenging to product

consistency, management of product performance, and engineering control of final product; also, further crystallization of these glass may introduce additional manufacturing processes as well as energy consumption. Hence, it is believed that amorphous glass foams are desirable for the development of this new melt-casting and hence float-manufacturable (can be manufactured as float glass) method. The given CaO-SnO2-P2O5-SiO2 glasses show minimum differences in stoichiometries between batched (see Table 1) and as-melted compositions (see Table 2). It can be challenging to study the chemistries of as-prepared

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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 36

glass foams if crystalline phases were present in the samples, i.e. the XPS will be detecting chemical information from the amorphous phase and crystalline phase simultaneously. Therefore, the absence of crystalline phases allows the study of correlations between glass network and physical properties of resultant glass foams. Microstructural and thermal characterization

Figure 2 Two-dimensional optical profile image of (a) Sn1 glass foam, (c) Sn2 glass foam, (e) Sn3 glass foam and (g) Sn4 glass foam; three-dimensional optical profile image of (b) Sn1 glass foam, (d) Sn2 glass foam, (f) Sn3 glass foam and (h) Sn4 glass foam.

Figure 2 shows the optical profile image (both two- and three-dimensional images) of the fracture surface of Sn1, Sn2, Sn3, and Sn4 samples. The three-dimensional images were processed with 3rd order levelling in order to remove the tilt resulted from the fracture surface. Such levelling process does not change the accuracy of measurement related to pore size as well as its distribution

15

. As shown in Figure 2, all

imaged pores were homogeneously-sized closed-pores with small pores occasionally being found in the walls between two adjacent pores. Pores were formed via the chemical reaction of ammonium diphosphate: negatively charged nitrogen ions were oxidized at elevated temperature to produce NOX gases, which then forms the porous structure found in this study. Generally, two types of porous structure were reported in glass foam studies 5: 1) open pores formed by the decomposition of carbonates, nitric

ACS Paragon Plus Environment

Page 9 of 36 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 Sustainable Chemistry & Engineering

acids, structural water, and sulfides; 2) close pores formed during a redox (mostly oxidation reactions) process of ammoniums, nitrides, carbides. Given the pores were closed-pores, it is believed that the formation of pores in this study is primarily associated with the ammonium diphosphate (redox type foaming agent). Therefore, the relationship between close porosity and “foaming agent (ammonium diphosphate)” in our work agrees with results being reported elsewhere 5. It should be pointed out that the emission of NOX glass would be minimal since the pores were closed. The average pore size (obtained from 10 images of each sample) of Sn1, Sn2, Sn3, and Sn4 is 99.9 ±3.3, 88.9 ±3.7, 78.5 ±2.4, and 60.9 ±2.2 µm, respectively. It was found that the average pore size of CaOSnO2-P2O5-SiO2 decreased with the addition of SiO2. The addition of SiO2 (a glass network former) increased cross-linking of the glass network, increased the viscosity of the glass forming liquids at elevated temperature

16-17

, and limited the motion of gas bubbles according to Stokes law (drag forces

increase with increasing viscosity)

18

. Given that the motion of gas bubbles can be limited with higher

viscosity, the forming of larger pores can be reduced since they are being formed by the merging of small gas bubbles. Therefore, the pore size of these melt-casted glass foams can be tuned via the addition of SiO2. Not only a tunable pore size of glass foam will help tailoring the final product to different applications, but also decrease the dimension of complexity during manufacturing processes.

Figure 3 DSC upscans of as-prepared glass foam samples. The scale of the DSC plots was adjusted, and first derivatives were used to identify glass transition as well as crystallization exothermic peaks.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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 effect of SiO2 addition on the network rigidity was evaluated using thermal properties including glass Tg (glass transition temperature) and CTE. As shown in Figure 3, the Tg increased with the addition of SiO2; the Tg of Sn1, Sn2, Sn3, and Sn4 is 462.6, 469.1, 476.9, and 483.6 oC, respectively (see Table 3). Tg and CTE can be used as measures of glass network rigidity, and higher Tg and CTE values often represent a higher network rigidity as well as more developed cross-linking between network formers (Si and P in this study) 19-20. The addition of SiO2 increased the network rigidity of glass foams, and further decreased the average pore size by increasing the viscosity at elevated temperature. Similar results on CTEs were found (see Table 3), and lowered CTEs are beneficial to the thermal shock resistance of glass foams 21-24. Table 3. Thermal properties of as-prepared glass foams Sample Sn1 Sn2 Sn3 Sn4

CTE (107 o / C) 84.6±1.1 81.2±0.8 78.3±0.6 74.6±1.3

Tg (oC) 462.6 469.1 476.9 483.6

The crystallizing events shown in Figure 3 were not studied given the fact that the crystallization temperature (above 700 oC) is much higher comparing to the annealing temperature (500 oC). During the processing of these glass foams, there were no crystallizing events as suggested by XRD results (see Figure 1). Correlation between glass networks and properties of foam glass Aforementioned glass compositions were designed to have a Sn/P ratio less than 0.25, which can keep the Sn-P network as a linear chain network

25-26

. A four-fold [PO4] unit has one non-bridging oxygen

termination and three bridging oxygen. Therefore, a Sn/P ratio less than the stoichiometry of linear chained Sn(n+2)/4PnO3n+1 kept the Sn-P network as linear chains, and further allows the addition of SiO2 as a network former. Therefore, Si and P were acting as glass network formers in all prepared samples while Ca and Sn were acting as network modifiers. Since the Sn-P network was designed to be linear chains, the

ACS Paragon Plus Environment

Page 10 of 36

Page 11 of 36 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 Sustainable Chemistry & Engineering

properties of as-prepared glass foams (e.g. network rigidity, network connectivity, thermal properties, and pore size) should hypothetically vary as a function of Si4+/Ca2+ ratio (oxidation state of Ca ion was determined using XPS). In order to investigate the relationship between the glass network and thermal properties, the ratio of Si4+/Ca2+ (addition to the Sn-P network) was quantitatively related to thermal properties.

Figure 4 Glass transition temperature as a function of Si4+/Ca2+. An R-square better than 0.99 was reported for the shown linear regression.

Figure 4 is the plot of Tg as a function of Si4+/Ca2+. Tg increased linearly (R2 >0.99) with the increase of Si4+/Ca2+, which increases the cross-linking between network forming units such as [PO4] and [SiO4]. Since the Sn-P network is linearly chained, the addition of [SiO4] can connect the Sn-P chains via Ca2+. Calcium ion, as a divalent ion, can be used to connect [PO4] chains (not fully charge compensated by Sn4+ since Sn/P ratio is less than 0.25) to [SiO4] units via the -O-Ca-O- bond. Therefore, the linear relationship between Tg and Si4+/Ca2+ can be attributed to the cross-linking between chained [PO4] and [SiO4] units via divalent modifying ion Ca2+ 27-28. A similar linear relationship (R2 >0.99) between CTE and Si4+/Ca2+ was found as well, as shown in Figure 5.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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

Figure 5 CTE as a function of Si4+/Ca2+. An R-square better than 0.99 was reported for the shown linear regression.

It needs to be pointed that thermal properties, CTE and Tg in particular, are closely related to the network connectivity and rigidity of glasses. In addition to these two properties, viscosity of glass forming liquids can reflect the network properties as well. In this study, the viscosity of glass forming liquids was related to the average pore of as-prepared glass foams since the motion of gas bubbles are determined by the viscosity of glass forming liquids. Figure 6 shows the relationship between average pore size and Si4+/Ca2+.

Figure 6 Pore size as a function of Si4+/Ca2+. An R-square better than 0.98 was reported for the shown linear regression.

ACS Paragon Plus Environment

Page 12 of 36

Page 13 of 36 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 Sustainable Chemistry & Engineering

As shown in Figure 6, the pore size decreases linearly when the ratio of Si4+/Ca2+ increases. As discussed previously, the increase of network connectivity or rigidity (e.g. Si4+/Ca2+) increases the viscosity of glass forming liquids. An increased viscosity or drag force limits the motion of small gas bubble (formed via redox of nitrogen in ammonium diphosphate) in the glass forming liquid, and therefore decreases the apparent pore size by limiting the merge of small gas bubbles. However, the area percentage porosity of the glass foams does not change (within statistical error) as a function of Si4+/Ca2+, as shown in Figure 7.

Figure 7 Plot of area percentage porosity and flexural strength as a function of Si4+/Ca2+. An R-square better than 0.98 was reported for the shown linear regression.

Within statistical error (STD is around 2 %), the area percentage porosity does not change as a function of Si4+/Ca2+. The volume of gas bubbles retained in each glass melt should be the same since the concentrations of ammonium diphosphate (foaming agent) from Sn1 to Sn4 remains identical. Therefore, the change of glass chemistry (Si4+/Ca2+) only alters the average pore size by changing the thermal properties of the glass forming liquids. On the other hand, the area percentage porosity is related to the amount of foaming agent added instead of being dependent on the glass chemistry. In addition to the linear relationship between glass network and thermal properties or pore size, the flexural strengths (a.k.a. modulus of rupture, MOR) were found to be linearly related to the ratio of Si4+/Ca2+ (see Figure 7) since the network rigidity is also related to strength of glasses. We conclude that the ratio of Si4+/Ca2+determines the glass network properties of this particular glass system and changes the thermal

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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 36

properties of as-prepared glass foams. Therefore, the tunability of CaO-SnO2-P2O5-SiO2 glass foam in thermal properties, pore size, and mechanical strength was achieved via the manipulation of glass network. Perspectives of cost-efficiency and potential impact This paper presents a series of novel glass foam that can be produced in a sheet-by-sheet manner similar to float glass. Conventional methods of manufacturing glass foams often involve melting and grinding of glass frits, molding of glass foams, and sintering of final products

1-3, 29

. The relative complicated

processes of conventional glass foams as well as the high energy consumption during those processes limited a cleaner and more efficient manufacturing of glass foams. In contrast, the method shown in this study presents melt-casted glass foams (mold-free and non-sintering) that are suitable to be produced in a highly automated glass float line in just one step. However, mechanical tools related to this melt-casting and float-manufacturable glass foams have yet to be studied and the trade-offs between cost of tin oxide and reduced processing temperature or complexity of processes need to be further studied. Apart from the manufacturing side of these CaO-SnO2-P2O5-SiO2 glasses, the unique combination of thermally tunable properties with controllable pore size made these glasses desirable for manufacturing processes. In addition, it was reported that Sn-M-P (M=alkaline earth, Mg, Zn, Ca, Sr, and Ba) glass system has lower viscosity and higher chemical durability than other low melting glass systems such as bismuth borosilicate, borates, tellurides, and lead-containing glasses 28, 30-31. A lowered viscosity helps the reduction of energy required for the melting and annealing processes of manufacturing foam glass since the liquidus temperature (or so called melting temperature) and Tg (glass transition temperature) are lowered

32-33

. Thus, the advantages of CaO-SnO2-P2O5-SiO2 glass foams presented in this study are

summarized as follows: 1. thermal properties and pore size can be tuned via the manipulation of glass network, i.e. connecting the linearly chained Sn-P network with [SiO4] units using Ca2+;

ACS Paragon Plus Environment

Page 15 of 36 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 Sustainable Chemistry & Engineering

2. the glass foams show potential to be manufactured in a sheet-by-sheet float glass line since it is a melt-casted glass foam; 3. the network of these glasses make themselves suitable as constructional materials with good thermal shock resistance (low CTE), insulation properties (close porosity), and chemical durability. Conclusions Float-manufacturable CaO-SnO2-P2O5-SiO2 glasses were prepared using a one-step melt-casting method. The closed porosity was created by the redox reaction of negatively charged nitrogen in ammonium diphosphate. The linearly chained Sn-P network was connected with [SiO4] units by Ca2+, and such increase in cross-linking increased the viscosity of glass forming liquids. Therefore, the NOx gas bubbles were retained in the melt to form porous structure. In addition, the manipulation of network via alkaline earth and silica makes the as-prepared glass foams tunable in their thermal properties, mechanical strengths and pore sizes. Glass foams in present study are promising in reducing the cost-efficiency of glass foam production, and revolutionize the processes in glass foam production by introducing a floatmanufacturable process.

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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

Acknowledgement YG thanks Dr. Carlo Pantano of the Pennsylvania State University for useful discussions. YG and JX thank RCB research group for the facilities used in this study. HG and CZ thank the financial support from Science and Technology R&D Program Focused on Special Project (No. 2017YFB0310201-02), International Science and Technology Cooperation Program of China (No. 51472151), Natural Science Foundation of China, Science and Technology Research Project of Xianyang (No. 2016K02-28), Shanxi Province R&D Project (No. 2017ZDXM-NY-048). Finally, we thank three anonymous reviewers for their effort in improving the quality of our article and scientific acumen. Author contributions YG, JX and RCB conceived the research. HG, CZ, MV, and YG performed the experimental work. HG performed the thermal analysis. YG and JX performed the XPS and other characterization work. CZ performed the four point bending test. YG and JX researched into the industry impact of this work. All authors discussed and analyzed the data. YG wrote the paper. All authors discussed and commented on the manuscript.

ACS Paragon Plus Environment

Page 16 of 36

Page 17 of 36 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 Sustainable Chemistry & Engineering

References 1.

Arulrajah, A.; Disfani, M. M.; Maghoolpilehrood, F.; Horpibulsuk, S.; Udonchai, A.; Imteaz, M.;

Du, Y.-J., Engineering and Environmental Properties of Foamed Recycled Glass as a Lightweight Engineering Material. J CLEAN PROD 2015, 94, 369-375, DOI 10.1016/j.jclepro.2015.01.080 2.

Gong, Y.; Dongol, R.; Yatongchai, C.; Wren, A.; Sundaram, S.; Mellott, N., Recycling of Waste

Amber Glass and Porcine Bone into Fast Sintered and High Strength Glass Foams. J CLEAN PROD 2016, 112, 4534-4539, DOI 10.1016/j.jclepro.2015.09.052 3.

Ponsot, I.; Bernardo, E., Self Glazed Glass Ceramic Foams from Metallurgical Slag and Recycled

Glass. J CLEAN PROD 2013, 59, 245-250, DOI 10.1016/j.jclepro.2013.06.029 4.

Guo, H.; Gong, Y.; Gao, S., Preparation of High Strength Foam Glass–Ceramics from Waste

Cathode Ray Tube. MATER LETT 2010, 64, 997-999, DOI 10.1016/j.matlet.2010.02.006 5.

Taurino, R.; Lancellotti, I.; Barbieri, L.; Leonelli, C., Glass–Ceramic Foams from Borosilicate

Glass Waste. INT J APPL GLASS SCI 2014, 5, 136-145, DOI 10.1111/ijag.12069 6.

König, J.; Petersen, R. R.; Yue, Y., Influence of the Glass–Calcium Carbonate Mixture's

Characteristics on the Foaming Process and the Properties of the Foam Glass. J EUR CERAM SOC 2014, 34, 1591-1598, DOI 10.1016/j.jeurceramsoc.2013.12.020 7.

D'Amore, G. K. O.; Caniato, M.; Travan, A.; Turco, G.; Marsich, L.; Ferluga, A.; Schmid, C.,

Innovative Thermal and Acoustic Insulation Foam from Recycled Waste Glass Powder. J CLEAN PROD 2017, 165, 1306-1315, DOI 10.1016/j.jclepro.2017.07.214 8.

Fernandes, H. R.; Ferreira, D. D.; Andreola, F.; Lancellotti, I.; Barbieri, L.; Ferreira, J. M.,

Environmental Friendly Management of Crt Glass by Foaming with Waste Egg Shells, Calcite or Dolomite. CERAM INT 2014, 40, 13371-13379, DOI 10.1016/j.ceramint.2014.05.053 9.

Fernandes, H.; Tulyaganov, D.; Ferreira, J., Preparation and Characterization of Foams from

Sheet Glass and Fly Ash Using Carbonates as Foaming Agents. CERAM INT 2009, 35, 229-235, DOI 10.1016/j.ceramint.2007.10.019

ACS Paragon Plus Environment

ACS Sustainable Chemistry & Engineering 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

10. Glasses

Page 18 of 36

Petersen, R. R.; König, J.; Yue, Y., The Mechanism of Foaming and Thermal Conductivity of Foamed

with

Mno2.

J

NON-CRYST

SOLIDS

2015,

425,

74-82,

DOI

10.1016/j.jnoncrysol.2015.05.030 11.

Petersen, R. R.; König, J.; Smedskjaer, M. M.; Yue, Y., Effect of Na2co3 as Foaming Agent on

Dynamics and Structure of Foam Glass Melts. J NON-CRYST SOLIDS 2014, 400, 1-5, DOI 10.1016/j.jnoncrysol.2014.04.029 12.

König, J.; Petersen, R. R.; Yue, Y., Fabrication of Highly Insulating Foam Glass Made from Crt

Panel Glass. CERAM INT 2015, 41, 9793-9800, DOI 10.1016/j.ceramint.2015.04.051 13.

Gong, Y.; Wren, A. W.; Mellott, N. P., Quantitative Morphological and Compositional

Evaluation of Laboratory Prepared Aluminoborosilicate Glass Surfaces. APPL SURF SCI 2015, 324, 594604, DOI 10.1016/j.apsusc.2014.10.132 14.

Apel, E.; Deubener, J.; Bernard, A.; Höland, M.; Müller, R.; Kappert, H.; Rheinberger, V.;

Höland, W., Phenomena and Mechanisms of Crack Propagation in Glass-Ceramics. J MECH BEHAV BIOMED MATER 2008, 1, 313-325, DOI 10.1016/j.jmbbm.2007.11.005 15.

Wieland, M.; Hänggi, P.; Hotz, W.; Textor, M.; Keller, B.; Spencer, N., Wavelength-Dependent

Measurement and Evaluation of Surface Topographies: Application of a New Concept of Window Roughness and Surface Transfer Function. WEAR 2000, 237, 231-252, DOI 10.1016/S00431648(99)00347-6 16.

Eichler, K.; Solow, G.; Otschik, P.; Schaffrath, W., Bas (Bao· Al2o3· Sio2)-Glasses for High

Temperature Applications. J EUR CERAM SOC 1999, 19, 1101-1104, DOI 10.1016/S09552219(98)00382-3 17.

Mauro, J. C.; Yue, Y.; Ellison, A. J.; Gupta, P. K.; Allan, D. C., Viscosity of Glass-Forming

Liquids. PROC NATL ACAD SCI 2009, 106, 19780-19784, DOI 10.1073/pnas.0911705106 18.

Ramos, J., Behavior of Multicomponent Gas Bubbles in Glass Melts. J AM CERAM SOC 1986,

69, 149-154, DOI 10.1111/j.1151-2916.1986.tb04720.x

ACS Paragon Plus Environment

Page 19 of 36 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 Sustainable Chemistry & Engineering

19.

Zhang, M.; Mancini, S.; Bresser, W.; Boolchand, P., Variation of Glass Transition Temperature,

Tg, with Average Coordination Number,> M M M