Mechanical, thermal, and structural investigations of chemically strengthened Na2O–CaO–Al2O3–SiO2 glasses

For a series of conventional soda-lime-silicate glasses with increasing Al₂O₃ content, we investigated the thermal, mechanical, and structural properties before and after K⁺-for-Na⁺ ion-exchange strengthening by exposure to molten KNO₃. The Al-for-Si replacement resulted in increased glass network polymerization and lowered compactness. The glass transition temperature (T_g), hardness (H) and reduced elastic modulus (E_r), of the pristine glasses enhanced monotonically for increasing Al₂O₃ content. H and E_r increased linearly up to a glass composition with roughly equal stoichiometric amounts of Na₂O and Al₂O₃ where a nonlinear dependence on Al₂O₃ was observed, whereas H and E_r of the chemically strengthened (CS) glasses revealed a strictly linear dependence. T_g, on the other hand, showed linear increase with Al-for-Si for pristine glasses while for the CS glasses a linear to nonlinear trend was observed. Solid-state ²⁷Al nuclear magnetic resonance (NMR) revealed the sole presence of AlO₄ groups in both the pristine and CS glasses. ²³Na NMR and wet-chemical analysis manifested that all Al-bearing glasses had a lower and near-constant K⁺-for-Na⁺ ion exchange ratio than the soda-lime-silicate glass. Differential thermal analysis of CS glasses revealed a “blurred” glass transition temperature (T_g) and an exothermic step below T_g; the latter stems from the relaxation of residual compressive stresses. The nanoindentation-derived hardness at low loads and <5 mol% Al₂O₃ showed evidence of stress relaxation for prolonged ion exchange treatment. The crack resistance is maximized for molar ratios n(M₍₂₎O)/n(Al₂O₃)$\approx \mathrm{1}$ for the CS glasses, which is attributed to an increased elastic energy recovery that is linked to the glass compactness.

1 Introduction

The readily tuned properties of glasses render them immensely useful for numerous applications (Shelby, 2005; Axinte, 2011; Zanotto and Mauro, 2017). However, a high brittleness is one disadvantage with glass (Hand and Tadjiev, 2010; Wondraczek et al., 2011), which is typically mitigated by increasing the glass thickness in many applications. Besides improved recycling procedures (Carr and Kim, 2017), a more sustainable utilization of thinner but stronger glasses reduces the amount of raw materials as well as the transportation costs of both precursors and products. There are several options for improvements, such as strengthening the glass and increasing its crack resistance (Wondraczek et al., 2011; Januchta et al., 2017; Karlsson and Wondraczek, 2021). For instance, the glass strength can be improved either by protective coatings or by introducing residual compressive stress (RCS) at the surface (Musil et al., 2002; Karlsson et al., 2010; Wondraczek et al., 2011; Varshneya et al., 2022). RCS may be implemented either by a conventional thermal tempering or by chemical strengthening (Wondraczek et al., 2011; Gross, 2019; Karlsson and Wondraczek, 2021).

Chemical strengthening is often referred to as ion-exchange strengthening because it typically involves replacing small (alkali metal) cation species by larger ones introduced from a molten salt (Wondraczek et al., 2011; Gross, 2019; Karlsson and Wondraczek, 2021), such as Li⁺ or Na⁺ substitutions by K⁺. The cation size-mismatch induces compressive stresses in the glass structure, which are subsequently relieved by a rapid relaxation process followed by a slower viscosity-dependent one (Varshneya, 2016; Macrelli et al., 2019). Therefore, the ion-exchange strengthening process is performed at a temperature around or below the strain point (a glass viscosity fixpoint, log₁₀ η = 14.5 dPa s). Some chemical strengthening procedures involve multiple consecutive ion-exchange processes, which are beneficial for producing an engineered stress profile (ESP) with a controlled RCS profile in the surface (Green et al., 1999). Chemically strengthened (CS) glasses are nowadays widely used for flat-screen covers in handheld electronic devices (Varshneya and Bihuniak, 2018), pharmaceutical cartridges (Morandotti and Zuccato, 2018), and for impact-resistant front windows in high-speed trains, aircrafts, and high-security windows (Sheikh et al., 2020); their utilization is expected to expand to various architectural (Zaccaria et al., 2021), automotive (Jacoby, 2018), solar (Allsopp et al., 2020), and flexible photonics (Macrelli et al., 2020) purposes.

Current CS glasses conform to the aluminosilicate (AS) or aluminoborosilicate glass systems (Gross, 2019; Karlsson and Wondraczek, 2021). Such “high-performance” CS glasses often incorporate large amounts of Al, which increase both the melting temperature and the melt viscosity (Silverman, 1939; Cheng et al., 2013; Karlsson, 2021), and thereby making the glasses comparatively costly to produce. The melt viscosity is maximized around a unity molar ratio of the metal oxide to alumina (Toplis et al., 1997; Webb et al., 2007), n(M₍₂₎O)/n(Al₂ O ₃) = 1 [henceforth referred to as “M₍₂₎O/Al₂O₃”]. Conventional soda-lime-silicate (SLS) glasses are in general unsuitable for chemical strengthening because they suffer from higher stress relaxation than high performance AS glasses (Varshneya and Kreski, 2012; Erdem et al., 2017; Güzel et al., 2019; Macrelli et al., 2019; Sun and Dugnani, 2020). Nonetheless, given that SLS glasses dominate the industrial market (Hand and Tadjiev, 2010; Deng et al., 2020), their minor doping with Al is a promising route towards an optimal balance between high chemical strengthening performance and lower production costs, and thereby more sustainable glasses (Wallenberger and Bingham, 2010; LaCourse, 2018).

To reach these goals, the structure of the glass and its properties must be contrasted before and after the chemical strengthening process. To explore and improve the understanding of the complex chemical strengthening mechanisms (Varshneya, 2010a; Macrelli et al., 2019; Karlsson and Wondraczek, 2021), we report herein on effects from K⁺-for-Na⁺ cation exchange on physical properties and local structures of the industrially relevant compositions 16Na₂O–12CaO–xAl₂O₃–(72–x)SiO₂ for increasing Al contents (x mol%). For both the pristine and ion-exchanged glasses, we discuss the trends of the glass transition temperature (T_g), the density (ρ) and its related molar volume (V_m) and atom packing density (compactness; C_g), crack resistance (CR) along with the nanoindentation-derived values of the hardness (H) and reduced elastic modulus (E_r). The degrees of K⁺-for-Na⁺ ion exchange was estimated by ²³Na magic-angle spinning (MAS) nuclear magnetic resonance (NMR) experiments. The Al speciation was probed by ²⁷Al MAS NMR, which revealed that all structures comprised essentially only [AlO₄]⁻ groups, regardless of the application of chemical strengthening.

2 Materials and methods

2.1 Glass preparation and composition analysis

Six glass samples were prepared with nominal compositions 16Na₂O–12CaO–xAl₂O₃–(72–x)SiO₂ with x = {0, 4, 8, 12, 16, 20} mol%; see Table 1. The glasses were melted in 200 g glass batches in Pt-Rh crucibles at 1500°C for x = {0, 4, 8} and 1600°C for x = {12, 16, 20} using Na₂SO₄ as a fining agent. The melts were quenched on a steel plate before being annealed at their respective calculated glass transition temperature. Full details of the glass sample preparation are given in ref. (Karlsson, 2021).

TABLE 1

TABLE 1. Analyzed and nominal glass compositions (in brackets), density (ρ), molar volume (V_m), compactness (C_g), oxygen packing density (C_O), glass transition temperature (T_g), average network connectivity (${\overline{\mathit{N}}}_{\mathbf{B}\mathbf{O}}^{\mathbf{S}\mathbf{i}}$), and the K⁺-for-Na⁺ exchange ratio^d,e (IE). Parameters based on nominal composition are given in parentheses. All data originate from the glass powders (see Section 2.2.1), except for density (ρ) and thus indirectly also molar volume (V_m), compactness (C_g) that originate from the glass monoliths (see Section 2.2.2).

Parts of the prepared glass were crushed on a steel plate using a steel hammer and then powdered using a mill from Retsch (model RM200) equipped with an agate mortar and pestle. The powdered glasses were sieved into two fractions, <45 µm and <90 μm, using steel sieves and a Retsch interval shaking table (model RV).

The glass compositions were determined by inductively coupled plasma optical emission spectroscopy (ICP OES) with an Aviro 200 instrument (Perkin Elmer). For each glass sample, 0.5 g of powdered material was dissolved in 15 ml 48% HF, 1 ml H₂SO₄ (95%–97%) and 0.5 ml HNO₃ (65%) altogether diluted to 50 ml using deionized water. The solution was evaporated, and the residual oxides were dissolved in 10 and 30 ml HCl and deionized H₂O respectively under mild heating on a hot plate for 60 min (occasionally deionized H₂O was added to replace evaporated H₂O). The solution was next transferred to a 250 ml beaker and 150 ml of deionized H₂O was added, whereupon the solution was evaporated to a volume of $\approx$100 ml. It was subsequently transferred to a volumetric flask and diluted to a total volume of 250 ml. The cation contents of the solutions were measured by ICP OES, whereas the corresponding O contents were determined from the charge requirements of the analyzed cations. Periodic Table Mix one for ICP measurements (SIGMA ALDRICH) was used as reference standards [33 elements in 10% HNO₃(aq)].

2.2 Chemical strengthening

2.2.1 Glass powders

About 35 g of reagent grade KNO₃ salt (ACROS Organics; 99% purity) and 5 g of the <45 µm sieved glass powder, q.v., Section 2.1, were heated to 450°C in separate fused quartz crucibles in a Nabertherm muffle furnace (model N2OH). The glass powder was then put into the molten KNO₃ and was stirred. One more stirring was performed after 2.5 h. After 5 h, the crucibles were taken out from the furnace and left to cool to ambient temperature on a refractory brick. Deionized water was added under magnetic stirring to dissolve the KNO₃, whereafter, the glass powder was separated from the aqueous solution using a Munktell 3-grade filter paper and a Büchner funnel.

2.2.2 Glass monoliths

For each prepared glass sample, two glass monoliths of approximate size 2 × 2 × 2 cm were cut out by a diamond saw. The glasses were polished in the Struers LaboForce100 equipment using the pertinent sample holder fixated by the UniForce equipment. SiC #320 sheets (MD Gekko) were used for rough polishing before fine polishing by the sheets MD Plan, MD Dac and MD Nap (in that order) with their corresponding polishing solutions. The samples were cleaned using Struers Lavamin equipment between the polishing steps.

The polished glass pieces were then subjected to ion-exchange strengthening in a Hybe muffle furnace (model KUT180) at 450°C. The glass pieces were heated in a wire net basket separately to a stainless-steel container with the salt. The salt bath consisted of 99% purity KNO₃ from ACROS Organics. The salt bath was first melted at 450°C before the basket with the glass pieces was put inside for 5 and 16 h. The basket was then taken out and the furnace was turned off with the door left a little open so that the cooling rate was ca. 3–5°C/min. The salt bath temperature was controlled to be within ±5°C to the target using a thermocouple of type K using a P655 logger from Dustmann Electronic GmbH. The treatment temperature, 450°C, was chosen specifically to give some stress relaxation (Svenson et al., 2016). The glass series from these treatments are called Al-CS 5 h and Al-CS 16 h.

2.3 Glass characterization

2.3.1 Solid state nuclear magnetic resonance experiments and data analysis

All ²³Na (spin 3/2) and ²⁷Al (spin 5/2) solid-state NMR experiments were performed with a Bruker Avance-III spectrometer at a magnetic field of 14.1 T, which provided the Larmor frequency of −158.7 MHz for ²³Na and −156.4 MHz for ²⁷Al. A fine powder (<10 µm particles) of each glass specimen was filled in a 3.2 mm zirconia rotor that underwent magic-angle spinning (MAS) at 24.00 kHz. To obtain NMR spectra that quantitative reflect the site populations (Edén, 2020), short radio-frequency (rf) pulses were employed: 0.40 μs (12° flip angle; 83 kHz nutation frequency) for ²³Na, and 0.32 μs (10° flip angle; 87 kHz nutation frequency) for ²⁷Al. The relaxation delays were 5 s (²³Na) and 2 s (²⁷Al), with 2024–10240 and 4048–8192 signal transients accumulated per ²³Na and ²⁷Al NMR acquisition, respectively, with the precise number depending on the Na and Al content of the glass. ²³Na and ²⁷Al shifts are quoted relative to 0.1 M NaCl(aq) and 1.0 M Al(NO₃)₃(aq), respectively.

Iterative numerical fitting of the MAS NMR spectra were used for determining the average values of the isotropic chemical shift, ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^E$, and quadrupolar product, ${\overline{C}}_{Q\eta }^E$, of each detected E = {²⁷Al, ²³Na} species in the respective ²⁷Al and ²³Na NMR spectra. ${\overline{C}}_{Q\eta }^E$ is the average quadrupolar product over all N ²⁷Al^j (or ²³Na^j) sites in the structure,

${\overline{C}}_{Q\eta }^E=\sqrt{\frac{1}{N}{\displaystyle \sum }_{j=1}^N{\left(C_Q^j\right)}^2\left(1+{\eta }_j^2/3\right)}(1)$

where $C_Q^j$ = e²q_jQ/h and $\eta$_j are the quadrupolar coupling constant and asymmetry parameter of the electric field gradient (EFG) tensor of ²⁷Al^j/²³Na^j, respectively (Edén, 2020). Each NMR spectrum was fitted by assuming one ²⁷Al or ²³Na nuclear site with NMR-parameter distributions (see Section 3.1). The best-fit parameters did not vary significantly if the central-transition (CT) peak was fitted alone or if the spinning sidebands of the satellite transitions were also included. The fitting was performed by in-lab developed software that minimizes the root-mean-square deviation between numerically simulated NMR spectra and the experimental counterpart while varying the targeted {${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^E,{\overline{C}}_{Q\eta }^E$, $W_{\mathrm{i}\mathrm{s}\mathrm{o}}^E$} parameter triplet; e.g., see (Iftekhar et al., 2012) for further information. Here, $W_{\mathrm{i}\mathrm{s}\mathrm{o}}^E$ is the width of the (as-assumed) Gaussian distribution of isotropic chemical shifts, whereas the Czjzek model accounted for the quadrupolar-product distribution (Czjzek et al., 1981).

The numerically exact simulations accounted for both first- and second-order quadrupolar interactions. The carousel-averaging COMPUTE algorithm (Levitt and Edén, 1998) was employed to efficiently exploit the time periodicity and for analytical averaging over the “γ” Euler angle, which together with a ^ROSELEBhemi18817 set of {α, β} orientations (Stevensson and Edén, 2006) ensured a fully converged powder average. The smallest time-integration step in the solution of the Schrödinger equation was 81 ns. The initial density and detection operators were ρ(t = 0) = I_x and I⁺, respectively.

2.3.2 Scattered light polariscope experiments

The surface RCS of the CS glass monoliths, σ_s, were determined using SCALP-05, a scattered light polariscope equipment from GlasStress Ltd. It is a well-described method to determine RCS in glass, see (Hödemann et al., 2016; Sun et al., 2019; Hödemann et al., 2020). The measurement depth was 0.75 mm, the stress optic coefficient (ΔC) was assumed to be constant, ΔC = 2.72 TPa⁻¹, and the refractive index (RI) was assumed to be 1.52 for x = {0, 4, 8, 12} and 1.53 for x = {16, 20} where x is the Al₂O₃ content; q.v., Section 2.1, and Table 1. The glass monoliths were cleaned with conventional window cleaning solution and dried clean with a wipe. A thin layer of Cargille immersion oil (liquid code 5040, RI = 1.5200 at 632.8 nm) was sprayed onto the sample. Three repeated measurements were performed for each sample and the average value was used. The stress was calculated from the gradient of the fitted optical retardation curve which starts from two to three pixels into the sample, i.e., ca. 10 µm, hence the errors of the determined stresses are estimated to be ±30%. In fact, σ_s is rarely overestimated with SCALP-05 (only for ultra-high σ_s) but for simplicity the error is displayed as both under- and overestimation.

2.3.3 Simultaneous thermal analysis experiments

The <45 µm glass powder (q.v., Section 2.1) and the CS glass powder (q.v., Section 2.2.1) were run in a simultaneous thermal analysis (STA), NETSZCH STA 409 PC/PG. Both the differential thermal analysis (DTA) and thermal gravimetric analysis (TGA) signals were recorded. Al₂O₃ crucibles were used for the measurements that were run under N₂-gas flow. The heating rate was 10 K/min up to 1200°C. An empty crucible was used for baseline correction. The glass transition temperature (T_g) was determined by using the Proteus software provided by NETSZCH using standard procedure where each T_g value was determined from the onset of the first endothermic step of the DTA curve (Zheng et al., 2019). The TGA analyses of the glass powders, available in the dataset; q.v., Karlsson et al. (2022), revealed no significant mass losses during the measurements.

2.3.4 Density, molar volume and compactness

Densities were measured using Archimedes method in deionized water (ρ_w = 0.998 g/cm³). The water temperature was kept within 21.5°C ± 0.5°C for the density measurements. Density values of the glass samples (ρ) were calculated using

$\mathit{\rho }=\frac{{\mathit{m}}_{\mathbf{d}}}{{\mathit{m}}_{\mathbf{d}}-{\mathit{m}}_{\mathbf{w}}}\cdot {\mathit{\rho }}_{\mathbf{w}}(2)$

where m_d is the weight of the dry glass sample, m_w the weight of the sample immersed in deionized water and ρ_w being the density for water. Sample weights varied between 6 and 10 g and the densities were estimated to be reproducible to ±0.003 g/cm³.

The molar volume of the glass sample was calculated using

${\mathit{V}}_{\mathbf{m}}=\frac{{\displaystyle \sum }_{\mathbf{i}}{\mathit{x}}_{\mathbf{i}}{\mathit{M}}_{\mathbf{i}}}{\mathit{\rho }}(3)$

where x_i is the molar fraction, M_i the molar mass of element i and ρ the density of the glass. The compactness (atom packing density) was calculated according to

${\mathit{C}}_{\mathbf{g}}=\frac{{\displaystyle \sum }_{\mathbf{i}}{\mathit{x}}_{\mathbf{i}}{\mathit{V}}_{\mathbf{i}}}{{\mathit{V}}_{\mathbf{m}}}(4)$

where V_i is the volume contribution from oxide i (i = M_xO_y) calculated from the ion radii, $r_{\mathrm{M}}$ and $r_{\mathrm{O}}$ (Rouxel, 2007), by $V_{\mathrm{i}}=\frac{\mathit{4}}{\mathit{3}}\pi N_{\mathrm{A}}\left(x{r}_{\mathrm{M}}^{\mathit{3}}+y{r}_{\mathrm{O}}^{\mathit{3}}\right)$, and $N_{\mathrm{A}}$ is Avogadro’s constant. The ionic radii {1.18, 1.12, 0.39, 0.26, 1.35} in Å were taken from (Shannon, 1976) with the coordination numbers (CN) {8, 8, 4, 4, 2} for the ions {Na⁺, Ca²⁺, Al³⁺, Si⁴⁺, O^2-} (Cormier and Neuville, 2004; Le Losq et al., 2014).

The oxygen packing density (C_O), which is the molar amount of oxygen per unit volume of the glass (Ray, 1974; Grammes et al., 2020), was calculated according to

${\mathit{C}}_{\mathbf{O}}=\frac{{\mathit{N}}_{\mathbf{A}}}{{\mathit{V}}_{\mathbf{m}}}{\displaystyle \sum }_{\mathbf{i}}{\mathit{x}}_{\mathbf{i}}{\mathit{n}}_{\mathbf{O}}^{\mathbf{i}}(5)$

where $n_{\mathrm{O}}^{\mathrm{i}}=\left\{\mathit{1},\mathit{1},\mathit{3},\mathit{2}\right\}$ is the stoichiometric coefficient of the amount of O in the respective oxide {Na₂O, CaO, Al₂O₃, SiO₂} component.

2.3.5 Nanoindentation experiments

Hardness (H) and the reduced elastic modulus (E_r) were determined by nanoindentation on polished glass samples before and after ion-exchange treatment using the Oliver-Pharr method (Oliver and Pharr, 1992). H is defined by the maximum applied load (F_m) divided by the projected contact area (A_p); $H=F_m/A_p$. A_p is calculated by a polynomial fitting, $A_p\left(h_c\right)={C_{\mathit{0}}h}_c^{\mathit{2}}+{C_{\mathit{1}}h}_c^{\mathit{1}}+{C_{\mathit{2}}h}_c^{\mathit{1}/\mathit{2}}+{C_{\mathit{3}}h}_c^{\mathit{1}/\mathit{4}}+\dots +{C_{\mathit{8}}h}_c^{\mathit{1}/\mathit{128}}$, where C_x is the indenter specific factors ($C_{\mathit{0}}=\mathit{24.56}$ for a perfect Berkovich tip) and h_c is the real contact depth. h_c takes into account the sink-in effect which is calculated by $h_c=h_m-\epsilon \frac{F_m}{S}$, where h_m is the maximum penetration depth, ε is a tip factor (ε = 0.72 for a Berkovich tip) and S is the stiffness as determined from the slope upon unloading, $S=\frac{\partial P}{\partial h}$. The reduced elastic modulus (E_r) is determined through $E_r=\frac{\sqrt{\pi }}{\mathit{2}\beta }\frac{S}{\sqrt{A_p}}$, where β is a geometrical tip factor (β = 1.034 for a Berkovich indenter).

Nanoindentation was performed using a NHT² instrument from Anton Paar with indentation loads 1, 5, 10, 15, 25, 50 and 75 mN. The number of indents was 40 for 1 mN load and 20 for the other load levels. The indentations had the following settings: acquisition rate 10 Hz, loading and unloading rate two times the max load per min (i.e., 2 mN/min for the max load of 1 mN etc.), holding time at max load 10 s, approach speed 4 μm/s and the stiffness threshold 500 μN/μm. The Poisson’s ratio (ν), that is affecting the determination of E_IT by $E_{\mathrm{I}\mathrm{T}}=\left(\mathit{1}-{\nu }^{\mathit{2}}\right)/\left(\frac{\mathit{1}}{E_{\mathrm{r}}}-\frac{\left(\mathit{1}-{\nu }_i^{\mathit{2}}\right)}{E_{\mathrm{i}}}\right)$, where E_i and ν_i are the elastic modulus and Poisson’s ratio of the indenter tip. E_IT is only reported on in the dataset; see Karlsson et al. (2022), was hypothesized to be constant, ν = 0.23, although ν can be slightly affected by the increasing Al₂O₃ content (Luo et al., 2016b; Pönitzsch et al., 2016; Ren et al., 2019) and affected by the chemical strengthening (Ryou et al., 2020). Collected nanoindentation data are available in the dataset (q.v., Karlsson et al., 2022) and for 75 mN are selected data given in Supplementary Table S1.

2.3.6 Microindentation experiments

The crack resistance (CR) was measured on polished glass monoliths before and after ion-exchange treatment using a Micro-Combi Tester (MCT) from CSM Instruments. The Vickers microindentations were used to produce radial crack statistics in the glass samples The CR method is described in (Sundberg et al., 2019), which follows the original procedure of (Kato et al., 2010). 15 indents were made for each load using a Vickers indenter. The least square method was used for fitting to the Weibull cumulative function,

$\mathit{P}\mathit{C}\mathit{I}=\mathbf{1}-{\mathit{e}}^{-{\left({\mathit{x}}_{\mathit{p}}/{\mathit{x}}_{\mathit{c}}\right)}^{\mathit{m}}}(6)$

where PCI is the probability of crack initiation, x_p is the load in N, x_c is the characteristic value in N and m the Weibull modulus. The CR is then defined as the load when the PCI is 50%. However, the Weibull cumulative function, x_c, can likewise be used as a relative measure of the CR, cf. previous paper (Karlsson, 2022). The Vickers indentations were made with an acquisition rate of 10 Hz using a linear loading with an (un)loading rate of two times the max load per min (i.e., 2N/min for the max load 1N etc.). The holding time at load was 15 s. The approach speed was 8 μm/min and the retract speed 16.6 μm/min. The contact force was 30 mN and the contact stiffness threshold 25,000 μN/μm. The indentations were made in the load range up to 30N and in the displacement range up to 1000 µm. All indentations, both nano and micro were performed at 23°C ± 2°C and relative humidity of 40% ± 10%. We estimated the error of the CR and x_c to be ±20% except for the samples with m > 10 where the error is lower (cf., Supplementary Table S2).

3 Results

3.1 Solid state nuclear magnetic resonance

3.1.1 ²⁷Al magic-angle spinning nuclear magnetic resonance

Owing to the low/moderate cation field strength of the Na⁺ and Ca²⁺ ions of the Na₂O–CaO–Al₂O₃–SiO₂ glasses, a traditional network model is expected to hold well for their structures, meaning that the AS glass network only involves [AlO_4/2]⁻ groups, surrounded by SiO_4/2 tetrahedra that accommodates all non-bridging oxygen (NBO) ions (Engelhardt and Michel, 1987; Edén, 2020). The sole presence of tetrahedrally coordinated Al species is verified by the ²⁷Al MAS NMR spectra presented in Figures 1A–C, which were recorded from the Al8, Al16, and Al20 glasses. The observed ²⁷Al shifts at the peak maxima, δ_max, remain in the 57–59 ppm range, and are consistent with AlO₄ groups (Engelhardt and Michel, 1987; Edén, 2020). The slight “tailing” of the ²⁷Al NMR peak shapes towards lower ²⁷Al shifts stems from distributions of ²⁷Al quadrupolar products, ${\overline{C}}_{Q\eta }^{\mathrm{A}\mathrm{l}}$ (Edén, 2020).

FIGURE 1

FIGURE 1. ²⁷Al NMR spectra recorded at 14.1 T and 24.00 kHz MAS from the Na₂O–CaO–(Al₂O₃)–SiO₂ glasses before (A–C) and after (D–F) exchange with K⁺ ions. The peak maximum (in ppm) is indicated at the top of each NMR spectrum.

For increasing Al₂O₃ content, the NMR-peak maxima in Figure 1 displace concurrently towards higher shifts. From the best-fit ²⁷Al NMR parameters listed in Table 2 follows that the observed peak displacements originate from a concomitantly increased (average) isotropic ²⁷Al chemical shift (${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{A}\mathrm{l}}$) when SiO₂ is replaced by Al₂O₃. This is most transparent from Figure 2A that plots both δ_max and ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{A}\mathrm{l}}$ against the Al₂O₃ content, revealing a linear relationship between ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{A}\mathrm{l}}$and n(Al₂O₃). Along with the increased isotropic ²⁷Al chemical shift, the quadrupolar products also manifest a concurrent minor increase, which accounts for roughly half of the net ∼2 ppm increase in the full-width-at-half-maximum height (fwhm) observed between the ²⁷Al NMR responses from the Al8 and Al20 glasses; the remaining ∼1 ppm peak-broadening stems from a very minor increased chemical-shift dispersion ($W_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{A}\mathrm{l}}$).

TABLE 2

TABLE 2. ²⁷Al and²³Na NMR parameters before and after chemical strengthening.^a

FIGURE 2

FIGURE 2. Average isotropic chemical shift (δ_iso) and the shift at the NMR-peak maximum (δ_max) plotted against the mol% of Al₂O₃ in the glass for (A) ²⁷Al and (B) ²³Na. “P” refers to results from the pristine glasses and “CS” to those obtained after chemical strengthening; see Table 2.

The negatively charged [AlO₄]⁻ moieties are expected to be charge-compensated primarily by the Na⁺ cations, whereas the remaining Na⁺ reservoir, along with that of Ca²⁺, mainly act as glass network modifiers that introduce—and charge balance—NBO anions at the SiO₄ groups (Mysen et al., 1981; Lee and Sung, 2008; Gambuzzi et al., 2014a). Consequently, the progressive Al₂O₃-for-SiO₂ substitutions (at constant Na₂O and CaO contents) are accompanied by a concurrently increased polymerization of the AS glass network. This feature is reflected by an increase of the average number of bridging oxygen (BO) atoms per SiO₄ group (${\overline{N}}_{\mathrm{B}\mathrm{O}}^{\mathrm{S}\mathrm{i}}$) (Edén, 2011), from ${\overline{N}}_{\mathrm{B}\mathrm{O}}^{\mathrm{S}\mathrm{i}}$ = 3.3 of the parent 16Na₂O–12CaO–72SiO₂ glass to ${\overline{N}}_{\mathrm{B}\mathrm{O}}^{\mathrm{S}\mathrm{i}}$ = 3.7 of the Al-richest sample of stoichiometry 16Na₂O–12CaO–20Al₂O₃–52SiO₂ (Table 1). The glass-network polymerization for increasing Al content is also reflected in enhanced physical glass properties, such as an increased glass transition temperature, hardness, and reduced elastic modulus (see Section 3.3 and Section 3.5). Such trends for increasing Al content are well documented (Cheng et al., 2013; Takahashi et al., 2015; Ragoen et al., 2017b).

Upon the K⁺-for-Na⁺ exchange, the ²⁷Al MAS NMR spectra (Figures 1D–F) reveal only minor alterations, which underlie the following NMR-parameter trends (Table 2): (i) Although the linear correlation between the isotropic chemical shift and n(Al₂O₃) remains intact after chemical strengthening, the ion-exchanged glass exhibits a 1–2 ppm lower ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{A}\mathrm{l}}$ value at a fixed Al content (Figure 2A), which is attributed to the partial K⁺-for-Na⁺ replacement, where K⁺ also charge compensate the [AlO₄]⁻ groups. (ii) Despite that the net fwhm value of the ²⁷Al MAS NMR peak remains essentially unaltered after chemical strengthening (Table 2), the quadrupolar product is slightly reduced by 0.25–0.5 MHz. Apparently, the accompanying NMR-peak narrowing is counter-balanced by a ∼0.7 ppm broadening from a slightly larger chemical-shift dispersion, thereby giving an overall unaltered ²⁷Al NMR peak-width (Table 2). Although the emphasized shift dispersion is consistent with the larger chemical/structural disorder created by mixed K⁺/Na⁺ cations in the second coordination sphere of the ²⁷Al^[4] nuclei, the decrease in ${\overline{C}}_{Q\eta }^{\mathrm{A}\mathrm{l}}$ is surprising and difficult to rationalize, as it suggests an overall enhanced local ordering around the ²⁷Al nuclei of the AlO₄ groups.

3.1.2 ²³Na magic-angle spinning nuclear magnetic resonance

We next consider the ²³Na MAS NMR results, where Figure 3A depicts the NMR spectrum from the SLS glass. As anticipated from the spin-3/2 nature of ²³Na, its resonance is broad and slightly tailed towards lower shifts (vide supra). On the introduction of Al in the glass network, the ²³Na NMR peak maximum shifts from −9.7 ppm (Figure 3A) to around −12.5 ppm in the spectra from the Al-bearing glasses Al8, Al16, and Al20; see Figures 3B–D and Table 2. This trend stems from an isotropic chemical shift decrease from ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{N}\mathrm{a}}=-$1.9 ppm in the SLS glass towards the asymptotic value ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{N}\mathrm{a}}=-$5.8 ppm for the two Al-richest glasses. This shift effect is attributed to the partial relocation of the Na⁺ cations from being associated with the Si–NBO moieties of the silicate glass to instead be positioned around the [AlO₄]⁻ tetrahedra. Note, however, that the slight ²³Na shift displacements of δ_max and ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{N}\mathrm{a}}$ may also stem from an altered distribution of Na coordination numbers.

FIGURE 3

FIGURE 3. ²³Na NMR spectra recorded at 14.1 T and 24.00 kHz MAS from the as-indicated glasses before (A–D) and after (E–H) exchange with K⁺ ions. Note that all NMR spectra are normalized to the same maximum peak height, and that the integrated signal intensity of each spectrum in (E–H) is only 25%–35% of that observed in the respective pristine glass shown in (A–D).

Relative to the pristine glasses, drastical ²³Na shift-changes are observed in the NMR spectra obtained from the series of K⁺-for-Na⁺ exchanged glasses displayed in Figures 3E–H, where the shift at the peak maximum is increased by 5–6 ppm (Table 2). As for ²⁷Al, this deshielding trend mirrors that of the isotropic ²³Na chemical shift (Figure 2B), which is increased by ∼6 ppm upon chemical strengthening. The same ²³Na deshielding trend has been reported for Na−Ca−Si−O glasses upon chemical strengthening by K⁺ incorporation (Ragoen et al., 2017a), and accords also with elevated-pressure induced effects observed in ²³Na MAS NMR from AS glasses (Lee, 2010).

The dependence of the isotropic ²³Na chemical shift on structural parameters is complicated but is known to increase for a decrease in either the Na coordination number (Na^[p]) or the average Na–O distance (Xue and Stebbins, 1993; George and Stebbins, 1995; Charpentier et al., 2004; Gambuzzi et al., 2014b; Yu et al., 2020), where we note that these structural parameters are usually correlated. Hence, the observed ²³Na shift increase upon the K⁺-for-Na⁺ exchange may stem from one or both of the following: (i) a shift in the distribution of Na^[p] average coordination towards lower p, where the larger K⁺ cation outcompetes Na⁺ for the highest coordination sites in the glass structure, or (ii) a net reduction in the average Na–O distances (with unaltered coordination numbers), as concluded by (Ragoen et al., 2018) or tacitly assumed (Ragoen et al., 2017a) from cation-exchanged Na/Ca-based silicate glasses. Moreover, unlike the scenario of Al, the partial replacement of Na⁺ cations by K⁺ leads to a minor increase in the ²³Na quadrupolar product by ∼0.4 MHz, the elevation of which remains constant, regardless of the presence/absence of Al in the glass (or the precise Al content); see Table 2.

From the integrated ²³Na NMR signal intensities of the pristine and ion-exchanged glasses in Figure 3 and the assumption of an equimolar K⁺-for-Na⁺ exchange, we estimated that a majority of the Na⁺ cations (74%) was exchanged in the Al-free SLS glass (Al0), whereas a somewhat reduced exchange efficiency (∼65%) resulted for the Al-bearing counterparts. Incidentally, the exchange-degree trend of an initial large drop at no/low Al₂O₃ contents, followed by near-constant values for increasing x ≥ 8 mol%, which match that of ${\overline{\delta }}_{\mathrm{i}\mathrm{s}\mathrm{o}}^{\mathrm{N}\mathrm{a}}$ in Figure 2B. The NMR-derived cation-exchange fractions agree very well with those estimated from the analyzed glass compositions (Table 1), but with a superior accuracy. The slightly lower cation exchange of the Al-rich glasses may be ascribed to their more polymerized—and thereby more rigid—AS glass networks, where the charge-balance requirement of the tetrahedrally coordinated Al to form [AlO₄]⁻$\cdots$Na⁺ pairs naturally reduce the Na⁺ mobility and thereby the cation exchange.

3.2 Residual compressive stress (RCS)

Figure 4 shows the RCS as a function of the Al₂O₃ content and the recorded σ_s values are given in Supplementary Table S1. The 5 and 16 h treatment do not follow the same trend; the 5 h follows a positive second degree polynomial function while the 16 h follows a negative second degree polynomial function. The RCS results will be further discussed in relation to other properties in the forthcoming sections.

FIGURE 4

FIGURE 4. Residual compressive stress (RCS) at surface, σ_s, as a function of Al₂O₃ content. The lines are polynomial fits of the second degree.

3.3 Glass transition temperature (T_g) and sub-T_g relaxation

Figure 5 displays the DTA data obtained from both the pristine and CS glass powders. All pristine samples exhibit a well-defined glass transition, with the T_g values increasing concurrently with the Al₂O₃ content, in good agreement with literature findings (Cheng et al., 2013; Takahashi et al., 2015; Ragoen et al., 2017b). The determined T_g for the investigated glasses are given in Table 1. The second endothermic peak, which we are not investigating in this paper, we anticipate being the onset of the crystallization range of the glasses.

FIGURE 5

FIGURE 5. DTA graphs for (A) Al0–20 glass powder and (B) CS Al0–20 glass powder. The red dashed lines are guides for the eye showing the trends.

The T_gs of the CS glasses are diffuse and not reported due to their large uncertainties. In contrast to the pristine glasses, the first step of the DTA curve is characterized by an exothermic step (hereinafter referred to as a sub-T_g step). We anticipate the observed sub-T_g step to stem from relaxation processes of the chemical-strengthening-induced RCS, analogous to the sub-T_g step observed from mechanically induced excess enthalpy studied by (Martin et al., 2005). The sub-T_g step coincides well to the point where stresses start to relax, i.e., slightly above the strain point. The Al8-CS glass manifests 2 T_g peaks, which is attributed to the presence of two separate segments/phases of each glass particle they presumably differ primarily in their glass modifier composition, constituting ion-exchange-induced K⁺-bearing outer segment and a core/bulk of the pristine Al8 glass (similar to a liquid-liquid phase separation (Zheng et al., 2020)). Such inhomogeneities rationalize the ill-defined T_g values of the CS glasses. Moreover, for the Al12-CS, Al16-CS and Al20-CS glasses, the endothermic steps are particularly diffuse, which is attributed to slower relaxation stemming from an increased viscosity due to the high Al content of these glasses (Karlsson, 2021). Al16-CS also exhibit a diffuse sub-T_g step. Apparently, the Al16-CS glass exhibits the highest temperatures for the sub-T_g step and T_g peak, as is shown by the dashed guiding lines in Figure 5.

The trends of Figure 5 correspond well to the recorded σ_s data shown in Figure 4 and given in Supplementary Table S1. Furthermore, the results of Figure 5 suggest that the relaxation of RCS might be investigated by differential scanning calorimetry (DSC) measurement on ion-exchanged glass powder or discs in analogy with the method of (Yue et al., 2002) for fictive-temperature estimates. However, more work is required to verify this concept.

3.4 Density, molar volume and compactness

The results of Table 1 reveal a concomitant increase in the glass density and molar volume when silica is replaced by alumina, in good agreement with previous findings (Cormier and Neuville, 2004; Ragoen et al., 2017b). These results may be understood from the free-volume expansion that accompanies the increased average Na–O bond length upon the Na–NBO–Si$\to$ Na–BO–Al bond replacements occurring when Na⁺ gradually shifts from balancing the NBO anions of silicate groups to BO atoms of [AlO₄]⁻ moieties (Cormier and Neuville, 2004; Webb et al., 2007; Mysen, 2021). It is thus reasonable that the free volume increase with increasing Al. In general, free volume should increase mobility, and is observed for Na⁺ self-diffusion coefficients (Frischat, 1975). However, for increasing Al content of the studied glasses, the exchange ratio (IE) is rather reduced; see Table 1. We emphasize that IE does not necessarily reveal the same trend as the effective diffusion coefficients as a function of Al₂O₃ but do give information about the effectiveness of the ion exchange, $C_{{\mathrm{K}}^{+}}/\left(C_{{\mathrm{K}}^{+}}+C_{{\mathrm{N}\mathrm{a}}^{+}}\right)$, where $C_{M^{+}}$ is the concentration of the alkali cation M⁺ in the surface. In the ternary system (Na₂O–Al₂O₃–SiO₂), additions of Al₂O₃ decreases the effective diffusion coefficients, though n.b., that Na₂O–SiO₂ give a less effective ion exchange in the surface (Ragoen et al., 2017b).

Along with an increase in the free volume, the C_g is decreased for increasing alumina content. The decrease of the C_g has also been observed in the ternary Na₂O–Al₂O₃–SiO₂ system (Bechgaard et al., 2016). C_g is linearly decreased for Al₂O₃ < 10 mol% and then shifting to a nonlinear trend for Al₂O₃ > 10 mol% which to some extent is caused by the deviation from the nominal composition, see Table 1. The C_g can affect other properties (Giri et al., 2004; Rouxel, 2007) and thus the structural moieties influencing the C_g may be a key for understanding the network dilation coefficient besides glass network connectivity (Tandia et al., 2012; Wang et al., 2017). For instance, the change of packing between Na and K (in alkali silicates) is the highest change for all the alkali ions as shown from the study of (Giri et al., 2004) and is a reason to why Na and K is most suitable for chemical strengthening compared to other alkali-pairs. The C_O remains approximately invariant when SiO₂ is replaced by Al₂O₃.

3.5 Hardness and reduced elastic modulus

Figure 6 shows H and E_r as a function of Al₂O₃ content for 75 mN load. For the pristine glasses, H and E_r increase with the Al₂O₃ content, at first linearly, but for M₍₂₎O/Al₂O₃ ≈ 1 nonlinearly. The results follow similar trends as reported previously (Yoshida et al., 2004). For the CS series, the trends are linear for both H and E_r. For Al₂O₃ < 5 mol% in the CS 16 h series have lower and approximately the same E_r as for the untreated series.

FIGURE 6

FIGURE 6. (A) Hardness (H) and (B) Reduced Elastic Modulus (E_r) for 75 mN nanoindentation as a function of Al₂O₃ content in mol%. The solid lines are polynomial fittings with the R² and polynomial degree given in the legend.

Figure 7 plots the hardness (H) against the indentation load for each glass before and after chemical strengthening. Generally, regardless of the precise load and treatment time, the chemical strengthening give an increase in H. However, the 16 h CS glasses typically give lower hardness values than the 5 h CS glasses in all cases except for M₍₂₎O/Al₂O₃ ≈ 1 (Table 1) where the 16 h give higher hardness than 5 h for the 5 and 15 mN loads. N.b., that at 1 mN loads and Al₂O₃ contents <10 mol%, H is lower for the 16 h CS glasses than for the pristine glasses (see further discussion in Section 4.1). In Supplementary Figure S1, the ΔH_CS as a function of load and compositions are shown, where ΔH_CS is given by H_CS–H_pristine (Karlsson, 2022). In general, the trends are linearly increasing with the Al₂O₃ content, the linear fitting R² for the 5–75 mN loads of the 5 and 16h CS glasses gives 0.87–0.97 and 0.85–0.94 respectively. For 1 mN load, the trend for ΔH_CS is entirely different; as a function of Al₂O₃ content, it initially shows a decrease and a clear minimum for Al8 before increasing linearly. The results will be further discussed in Section 4.1.

FIGURE 7

FIGURE 7. Hardness (H) as a function of indentation load for each sample, pristine and CS-glasses for 5 and 16h respectively. The solid lines are guides for the eye.

3.6 Crack resistance

Figure 8 shows the dependence of CR on the Al₂O₃ content. The pristine series show a marginal decrease with increasing Al₂O₃ concentration, as previously observed for Na₂O–Al₂O₃–SiO₂ glasses (Yoshida et al., 2004). For the 5 h CS glasses, there is a clear increase from Al₂O₃ > 8 mol% to a maximum for M₍₂₎O/Al₂O₃ ≈ 1; see Table 1. The 16h CS glasses also show a maximum for M₍₂₎O/Al₂O₃ ≈ 1. Both 5 and 16 h CS glasses show a decrease for Al20. The results for Al0 are lower than reported for a conventional SLS float glass (Sundberg et al., 2019), which could be explained by higher surface quality of float glass (Güzel et al., 2020) or sharpness of tip (Gross, 2012). CR for the Al-bearing glasses are similar to what has been reported by (Morozumi et al., 2015). The CR including the Weibull fitting results are given in Supplementary Table S2. The results and its explanation will be discussed in Section 4.2.

FIGURE 8

FIGURE 8. Crack resistance (CR) as a function of Al₂O₃ content. The insets show some selected characteristic microscope light images of indents. The solid lines are guides for the eye.

4 Discussion

In general, AS glasses frequently exhibit a turning point for properties M₍₂₎O/Al₂O₃ ≈ 1. Especially transport properties, such as melt viscosity, exhibit clear maxima in both the ternary and the quaternary Na₂O–(CaO)–Al₂O₃–SiO₂ systems (Toplis et al., 1997; Webb et al., 2007; Karlsson, 2021). However, other properties conventionally also give a trend-shift at M₍₂₎O/Al₂O₃ ≈ 1 (Smedskjaer et al., 2013; Cormier, 2021). This originates from the preference for the alkali ions to act as charge compensators of the negatively charged AlO₄ groups; q.v., Section 3.1.1. This has now also been shown to account for effects for CS glasses as shown in Section 3.2, Section 3.3, Section 3.5, and Section 3.6. Furthermore, trends in physical properties as a function of the ${\overline{N}}_{\mathrm{B}\mathrm{O}}^{\mathrm{S}\mathrm{i}}$ generally shows same behavior as for against Al₂O₃ content, i.e., deviations from linearity around the M₍₂₎O/Al₂O₃ ≈ 1 for both pristine and CS glasses. In Section 4.1, respectively Section 4.2 will primarily the causes of the ΔH_CS and the CR results be discussed.

4.1 Hardness, reduced elastic modulus and residual compressive stress

For a progressing K⁺-for-Na⁺ ion exchange, the glass densifies while compressive stresses accumulate at the surface (Luo et al., 2016a); q.v. Figure 4. Along with hot compression, H is often increased by that kind of densification (Svenson et al., 2017). Therefore, it is reasonable that the hardness increases upon chemical strengthening treatment which thus is linked to the RCS (Jang, 2009), as has previously been indicated by (Svenson et al., 2016) and supported by the correlation between σ_s (c.f., Supplementary Table S1) and ΔH_CS in Figure 9A. The σ_s data of 5h is likely less accurate than 16h due to the lower depth-of-layer and it is reflected in the fit. σ_s as a function of ΔE_r,CS give a slightly better fit, R² 0.37 and 0.98 for 5 and 16h respectively. N.b., that RCS is not of the same nature as for thermally strengthened glass (Kese et al., 2004). Free volume and thus also C_g (q.v., Section 3.4) is linked to ν (Rouxel, 2007) which is a governing property for the ability to densify; lower C_g gives higher densification-ability; see Figure 4 in (Rouxel et al., 2008), and related text for detailed reasoning. The support for this is manifested by Figure 9B, where σ_s as a function of C_g and even C_O give a fair correlation. The lack of perfect correlation is wrecked by a combination of the σ_s error and the complexity of the compressive stress buildup and relaxation, which is a function of other properties as well (Dugnani, 2020), thus more studies on this are needed.

FIGURE 9

FIGURE 9. Residual compressive stress (RCS) at surface, σ_s, as a function of (A) ΔH_CS and (B) C_g as well as C_O.

Figure 10A plots ΔH_CS and ΔE_r,CS against C_g. The trend is quite linear until it turns after M₍₂₎O)/Al₂O₃ ≈ 1 (Table 1). In contrast to a recent report on the Na₂O–CaO–TiO₂–SiO₂ system (Karlsson, 2022), the present Na₂O–CaO–Al₂O₃–SiO₂ glasses reveal nearly linear correlations between V_m and the ratio E_r/H, cf. Supplementary Figure S2. E_r/H is essentially a measure of the elastic volume recovery which will be discussed more in Section 4.2 (Rouxel, 2015). Similar correlation as for Figure 9A can be found for σ_s as a function of E_r/H as well.

FIGURE 10

FIGURE 10. (A) ΔH_CS, ΔE_r,CS and (B) Crack Resistance (CR) as a function of the compactness (C_g). The solid lines are guides for the eye except for Al-CS 5 h and Al-CS 16 h in (B) which are linear fittings.

The mixed alkali effect (MAE) is well-known to primarily affect cation-transport-related properties, but has also been shown to influence mechanical properties (Kjeldsen et al., 2014; Aakermann et al., 2015; Grund Bäck et al., 2019); for n(K₂O)/[n(Na₂O)+ n(K₂O)]≈1, both H and E is lower than for ratios n(K₂O)/[n(Na₂O)+ n(K₂O)] close to 0 (Kjeldsen et al., 2014; Weigel et al., 2016). This applies also to isostatically compressed glasses (Aakermann et al., 2015). For CS glasses, the MAE is not manifested in the same way as for as-melted glass (Varshneya, 2010a).

For our CS glasses the origin of ΔH_CS cannot be caused by the MAE, instead it must be caused by RCS and its accompanied densification. However, at 1 mN loads (q.v., Figure 6 and cf. Supplementary Figure S1) notable trend-breakers can be observed where ΔH_CS are negative for Al0, Al4 and Al8. We hypothesize this, and the fact that 16h CS glasses generally show lower ΔH_CS, to be caused by viscosity driven stress relaxation. Thus, H then resembles the K₂O-rich end of the MAE of H (i.e., giving a lower value than for pristine glass (Kjeldsen et al., 2014; Aakermann et al., 2015). Thus, for relaxed CS glasses we believe MAE plays a role. We hope to confirm this reasoning in a future study.

4.2 Crack resistance

Upon application of an indenter on a glass surface, the energy is dissipated by elastic and plastic deformation (Yoshida, 2019; Rouxel et al., 2021; Varshneya et al., 2022). These processes differ in that when the load is removed, the elastic deformation is completely reversed, whereas the plastic deformation leaves a permanent imprint. The energy dissipation by plastic deformation can be categorized into two processes: densification, and shear flow (Yoshida, 2019; Rouxel et al., 2021; Varshneya et al., 2022). At sufficiently high indenter force the ability of the glass to deform elastically and plastically is not sufficient, instead the energy is dissipated as cracks. These cracks can either form during loading or during unloading (Yoshida, 2019; Rouxel et al., 2021; Varshneya et al., 2022). The crack type depends on the glass composition, but brittle materials exhibit most frequently radial and half-penny cracking (Yoshida, 2019; Rouxel et al., 2021), i.e., cracks originating from the corners of the Vickers indenter.

In CS glasses, (Morris et al., 2004), showed that the radial crack initiation is inhibited by RCS, but suggested that lateral cracking, i.e., sub-surface cracks parallel to the surface; see Figure 11, is enhanced by the RCS. However, Morris et al. stated that lateral cracks are not as critical for the glass-strength degradation as the radial counterparts, at least not until they reach the surface. Lateral and Median cracking are sub-surface cracking (Cook and Pharr, 1990), and are not recorded in the CR test. This suggests that RCS, along with the accompanied densification gradient, enhances the CR by obstructing sub-surface cracking to reach the surface. Chipping, defined as sub-surface cracking reaching the surface and is manifested by giving a crater; see Figure 11, then occurs at sufficiently high load, q.v., Figure 8 insets for 10N of Al16-CS and Al20-CS. Chipping is then taken to imply PCI = 100% if it affects all sides of the indent. Chipping was clearly observed for Al₂O₃ contents >10 mol%. This reasoning is confirmed by a recent study by (Kang et al., 2020) where they observed median crack initiation upon indentation, likely beneath the RCS zone. However, their Al₂O₃ content is not mentioned in the paper. Xiaoyu et al. concluded that radial cracks are more easily initiated in CS AS glass with about 5 wt% Al₂O₃ (Xiaoyu et al., 2018), which is in the range of 3–4 mol%; see Table 1, and thus agrees with our results.

FIGURE 11

FIGURE 11. Schematics of the chipping process during unloading of the indenter, where the sub-surface lateral cracks grow and reach the surface.

As already stated, chipping was for Al₂O₃ > 10 mol% frequently observed to occur during unloading. In thin glass central tensile stresses from the balancing RCS could potentially induce such behavior, however, our monoliths are quite thick (ca. 2–2.5 cm), so we expect the central tension to be marginal. Instead, the clear difference in CR is governed by the Al₂O₃ content. Al₂O₃ > 10 mol% reduces the “fast” network relaxation by having a favorable structure, i.e., low alkali-induced NBO structure, a high reduced elastic modulus and lower C_g (see Figure 10B). We already implied in Section 4.1, that lower C_g is linked to a higher ability for densification (Rouxel et al., 2008; Varshneya et al., 2015) and thus a higher ability for RCS buildup which also increases the CR; q.v., σ_s data in Supplementary Table S1. In Figures 10A,B linear relation between CR and C_g is shown, R² is 0.98 and 0.92 for 5 and 16h respectively. In addition, viscous relaxation of the RCS is also inhibited by the higher viscosity that the Al₂O₃ content gives (Karlsson, 2021).

Varshneya previously described ion-exchange strengthening to be analogous to indentation (Varshneya, 2010b; a), having both elastic and plastic processes that governs the buildup of RCS. It was predicted that the compositions that exhibit less indentation depth and thus those having higher elastic modulus should perform better as CS glass. For our CS glasses the indentation depth at max load, h_m, as a function of the Al₂O₃ content is decreased; q.v., Figure 12, which supports the prediction of Varshneya. The indentation depth, h_m, follows linear but opposite trends as H as a function of Al₂O₃ content. In a previous paper, a rough correlation between the ratio E_r/H and CR was found (Karlsson, 2022). It is more frequently used in its inverted form, H/E_r, where materials having H/E_r >0.1 exhibit excellent tribological properties (Musil et al., 2002; Calahoo et al., 2016; Varshneya et al., 2022). H/E_r can be directly linked to the elastic recovery energy (U_E/U_T) in the indentation process by a proportionality factor, κ, which can be approximated to 5.17 in the range 0.08–0.12 for H/E_r (Yetna N'Jock et al., 2016). U_E/U_T upon indentation is given by

$\frac{{\mathit{U}}_{\mathit{E}}}{{\mathit{U}}_{\mathit{T}}}=\mathit{\kappa }\frac{\mathit{H}}{{\mathit{E}}_{\mathit{r}}}(7)$

and the calculated values for U_E/U_T are given in Supplementary Table S1 together with H³/E_r², which instead is a measure of the resistance to plastic deformation (Calahoo et al., 2016). N.b., that we assume the same κ also for those samples having H/E_r > 0.12. Both U_E/U_T and H³/E_r² show similar trends as H and h_m as a function of Al₂O₃ content, cf. Supplementary Figure S3. Thus, U_E/U_T and H³/E_r² also show similar trends as a function of CR; see Supplementary Figure S3. Thus, H and E_r are keys for understanding CR through the previously described relations. Indentation induced cracking is governed by mismatching stresses upon loading or unloading and a better balance between elastic and plastic dissipation should give higher CR. For the studied glasses it occurs when U_E/U_T >57%. An analysis of the indentation stress fields, and its effect of RCS in CS glasses would be useful to elucidate more on this.

FIGURE 12

FIGURE 12. (A) Indentation depth and (B) U_E/U_T as a function of Al₂O₃ content. The data originates from nanoindentations of 75 mN load; see Supplementary Table S1.

The crack initiation mode, radial or median, can be dependent on the composition apart from the RCS. For Al₂O₃ < 10 mol% there is sub-surface cracking but also much easier radial cracking. However, the CR test only considers the radial crack statistics and is thus unfavorable for the CS glasses with Al₂O₃ < 10 mol%, as radial cracks are still easily formed (like isostatically compressed glasses (Zhang et al., 1995; Svenson et al., 2014)). The radial crack lengths for the CS glasses are clearly shorter, q.v. Figure 8, consistent with observations in borosilicate CS glasses (Talimian and Sglavo, 2017). However, this measure was not considered in the present study.

5 Summary and conclusion

Our study on Al-doped soda-lime-silicate glasses upon K⁺-for-Na⁺ exchange manifested a complex property-dependence on their relative Si and Al contents. T_g, H and E_r of the pristine glasses were enhanced monotonically for increasing Al₂O₃ content. H and E_r showed for M₍₂₎O/Al₂O₃ ≈ 1 nonlinear trends, whereas the CS counterparts were found to be strictly linearly increasing with the Al-for-Si replacement. T_g and σ_s on the other hand, was linear for the pristine glasses and linear to nonlinear for the CS glasses as a function of Al₂O₃ content. Several physical properties, along with the silicate network connectivity, of both series of pristine and chemically strengthened glasses depended roughly linearly on their alumina content, except for clear deviations observed for glasses around the M₍₂₎O/Al₂O₃ ≈ 1 composition.

²⁷Al MAS NMR experiments revealed a dominance of [AlO₄]⁻ groups throughout all pristine and ion-exchanged glass structures, with the exchange manifested only by a minor ²⁷Al deshielding upon the replacement of Na⁺ by the less electronegative K⁺ cations. Significant alterations were observed in the ²³Na shifts observed by ²³Na MAS NMR, which presumably reflects an overall lowering of the (average) Na coordination numbers because the larger K⁺ cations outcompete those of Na⁺ for the highest-coordination sites in the glass structure. The extent of cation exchange was also estimated by ²³Na MAS NMR, revealing a markedly higher degree of exchange (≈74%) of the Al-free SLS glass relative to the Al-bearing glasses, all of which revealed an essentially identical exchange level (∼65%) of the Na⁺ cations; this observation is attributed to the more polymerized networks of the Al-bearing glasses and the preference for the monovalent Na⁺ cation to balance the [AlO₄]⁻ groups, which naturally impedes its mobility and thereby exchange by K⁺.

Differential thermal analysis data revealed that the ion-exchanged glasses exhibit an exothermic step just below a blurred T_g. We propose that these characteristics stem from a combination of stress relaxation of the residual compressive stresses induced by the K⁺-for-Na⁺ ion exchange.

The chemical strengthening consistently increased the hardness, except for glasses incorporating <10 mol% Al₂O₃ for the specific conditions of an 16 h exchange period and an applied load of 1 mN, which resulted in a lower hardness than that of the pristine glass. We attribute these observations to a combination of stress relaxation and the mixed alkali effect. Stress relaxation is also manifested by the fact that the 16 h treatment time consistently yielded lower H and E_r values relative to glasses treated for 5 h. These effects were less obvious for the Al-richer glasses with n(M₍₂₎O)/n(Al₂O₃) ≈ 1 and on the σ_s results regardless of composition.

The crack resistance of the pristine glasses becomes slightly lower with increasing Al₂O₃. The chemical strengthening has a pronounced enhancement on the crack resistance for glasses with n(M₍₂₎O)/n(Al₂O₃) ≈ 1. The chemical strengthening increases the elastic energy of recovery and the resistance to plastic deformation. The observed linear correlation between the crack resistance and the compactness suggests that C_g is a key parameter for the buildup of compressive stresses in the vitreous structure. Thus, our results suggest that the crack resistance (CR) of a chemically strengthened glass is a property governed by both the composition and structure, as well as residual compressive stress, where the CR is favored by a polymerization increase and compactness decrease; the latter is also linked to the residual compressive stress increase.

Data availability statement

The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation. The datasets generated for this study can be found in the Swedish Service Repository https://snd.gu.se/en. DOI: https://doi.org/10.5878/kp36-vd61.

Author contributions

SK: conceptualization, methodology, validity, formal analysis, investigation, resources, data curation, writing ‐ original draft, writing ‐ review and editing, visualization, project administration, funding acquisition. RM: investigation, data curation, writing ‐ review and editing. SA: investigation, data curation, writing ‐ review and editing. MP: methodology, formal analysis, investigation, data curation. JA: validity, resources, writing ‐ review and editing. BS: formal analysis. ME: methodology, validity, formal analysis, investigation, resources, writing ‐ original draft, writing ‐ review and editing, visualization.

Funding

SK acknowledges funding from FORMAS, the Swedish Research Council for Sustainable Development, Grant No. 2018-00707.

Acknowledgments

Ingemar Malmros at RISE Research Institutes of Sweden and Charlotte Parsland at Linnaeus University are greatly acknowledged for their help with dissolution of glass powder and ICP-OES measurements, respectively. We also acknowledge fruitful discussions with Prof. Yuanzheng Yue at Aalborg University, in relation to the DTA analysis.

Conflict of interest

Authors MP and JA were employed by GlasStress Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmats.2022.953759/full#supplementary-material

References

Aakermann, K. G., Januchta, K., Pedersen, J. A. L., Svenson, M. N., Rzoska, S. J., Bockowski, M., et al. (2015). Indentation deformation mechanism of isostatically compressed mixed alkali aluminosilicate glasses. J. Non-Cryst. Solids 426, 175–183. doi:10.1016/j.jnoncrysol.2015.06.028

CrossRef Full Text | Google Scholar

Allsopp, B. L., Orman, R., Johnson, S. R., Baistow, I., Sanderson, G., Sundberg, P., et al. (2020). Towards improved cover glasses for photovoltaic devices. Prog. Photovolt. Res. Appl. 28 (11), 1187–1206. doi:10.1002/pip.3334

CrossRef Full Text | Google Scholar

Axinte, E. (2011). Glasses as engineering materials: A review. Mater. Des. 32 (4), 1717–1732. doi:10.1016/j.matdes.2010.11.057

CrossRef Full Text | Google Scholar

Bechgaard, T. K., Goel, A., Youngman, R. E., Mauro, J. C., Rzoska, S. J., Bockowski, M., et al. (2016). Structure and mechanical properties of compressed sodium aluminosilicate glasses: Role of non-bridging oxygens. J. Non-Cryst. Solids 441, 49–57. doi:10.1016/j.jnoncrysol.2016.03.011

CrossRef Full Text | Google Scholar

Calahoo, C., Zhang, X., and Zwanziger, J. W. (2016). Nanoindentation study of the surface of ion-exchanged lithium silicate glass. J. Phys. Chem. C 120 (10), 5585–5598. doi:10.1021/acs.jpcc.6b00733

CrossRef Full Text | Google Scholar

Carr, A. S., and Kim, J. (2017). Closed loop supply chain for glass recycling: Current practices and challenges. Int. J. Integr. Supply Manag. 11 (4), 316–331. doi:10.1504/ijism.2017.089845

CrossRef Full Text | Google Scholar

Charpentier, T., Ispas, S., Profeta, M., Mauri, F., and Pickard, C. J. (2004). First-Principles calculation of ¹⁷O, ²⁹Si, and ²³Na NMR spectra of sodium silicate crystals and glasses. J. Phys. Chem. B 108 (13), 4147–4161. doi:10.1021/jp0367225

CrossRef Full Text | Google Scholar

Cheng, J., Xiao, Z., Yang, K., and Wu, H. (2013). Viscosity, fragility and structure of Na₂O–CaO–Al₂O₃–SiO₂ glasses of increasing Al/Si ratio. Ceram. Int. 39 (4), 4055–4062. doi:10.1016/j.ceramint.2012.10.258

CrossRef Full Text | Google Scholar

Cook, R. F., and Pharr, G. M. (1990). Direct observation and analysis of indentation cracking in glasses and Ceramics. J. Am. Ceram. Soc. 73 (4), 787–817. doi:10.1111/j.1151-2916.1990.tb05119.x

CrossRef Full Text | Google Scholar

Cormier, L. (2021). “Glasses: Aluminosilicates,” in Encyclopedia of materials: Technical Ceramics and glasses. Editor M. Pomeroy (Oxford: Elsevier), 496–518. doi:10.1016/B978-0-12-818542-1.00076-X

CrossRef Full Text | Google Scholar

Cormier, L., and Neuville, D. R. (2004). Ca and Na environments in Na₂O–CaO–Al₂O₃–SiO₂ glasses: Influence of cation mixing and cation-network interactions. Chem. Geol. 213 (1), 103–113. doi:10.1016/j.chemgeo.2004.08.049

CrossRef Full Text | Google Scholar

Czjzek, G., Fink, J., Götz, F., Schmidt, H., Coey, J. M. D., Rebouillat, J. P., et al. (1981). Atomic coordination and the distribution of electric field gradients in amorphous solids. Phys. Rev. B 23 (6), 2513–2530. doi:10.1103/PhysRevB.23.2513

CrossRef Full Text | Google Scholar

Deng, W., Spathi, C., Coulbeck, T., Erhan, K., Backhouse, D., Marshall, M., et al. (2020). Exploratory research in alternative raw material sources and reformulation for industrial soda-lime-silica glass batches. Int. J. Appl. Glass Sci. 11 (2), 340–356. doi:10.1111/ijag.14775

CrossRef Full Text | Google Scholar

Dugnani, R. (2020). Closed-form solution to the residual stresses in ion-exchanged silicate glass including concentration-dependent material properties. J. Non-Cryst. Solids 536, 120012. doi:10.1016/j.jnoncrysol.2020.120012

CrossRef Full Text | Google Scholar

Edén, M. (2011). The split network analysis for exploring composition–structure correlations in multi-component glasses: I. Rationalizing bioactivity-composition trends of bioglasses. J. Non-Cryst. Solids 357 (6), 1595–1602. doi:10.1016/j.jnoncrysol.2010.11.098

CrossRef Full Text | Google Scholar

Edén, M. (2020). Update on ²⁷Al NMR studies of aluminosilicate glasses. Annu. Rep. NMR Spectrosc. 101, 285–410. doi:10.1016/bs.arnmr.2020.07.002

CrossRef Full Text | Google Scholar

Engelhardt, G., and Michel, D. (1987). High-resolution solid-state NMR of silicates and zeolites. Chichester, New York: John Wiley & Sons.

Google Scholar

Erdem, İ., Guldiren, D., and Aydin, S. (2017). Chemical tempering of soda lime silicate glasses by ion exchange process for the improvement of surface and bulk mechanical strength. J. Non-Cryst. Solids 473, 170–178. doi:10.1016/j.jnoncrysol.2017.08.010

CrossRef Full Text | Google Scholar

Frischat, G. H. (1975). Ionic diffusion in oxide glasses. Aedermannsdorf, Switzerland: Trans Tech S.A.

Google Scholar

Gambuzzi, E., Charpentier, T., Menziani, M. C., and Pedone, A. (2014a). Computational interpretation of ²³Na mqmas NMR spectra: A comprehensive investigation of the Na environment in silicate glasses. Chem. Phys. Lett. 612, 56–61. doi:10.1016/j.cplett.2014.08.004

CrossRef Full Text | Google Scholar

Gambuzzi, E., Pedone, A., Menziani, M. C., Angeli, F., Caurant, D., and Charpentier, T. (2014b). Probing silicon and aluminium chemical environments in silicate and aluminosilicate glasses by solid state NMR spectroscopy and accurate first-principles calculations. Geochim. Cosmochim. Acta 125, 170–185. doi:10.1016/j.gca.2013.10.025

CrossRef Full Text | Google Scholar

George, A. M., and Stebbins, J. F. (1995). High-temperature ²³Na MAS NMR data for albite: Comparison to chemical-shift models. Am. Mineralogist 80 (9-10), 878–884. doi:10.2138/am-1995-9-1002

CrossRef Full Text | Google Scholar

Giri, S., Gaebler, C., Helmus, J., Affatigato, M., Feller, S., and Kodama, M. (2004). A general study of packing in oxide glass systems containing alkali. J. Non-Cryst. Solids 347 (1), 87–92. doi:10.1016/j.jnoncrysol.2004.08.103

CrossRef Full Text | Google Scholar

Grammes, T., Limbach, R., Bruns, S., van Wüllen, L., de Ligny, D., Kamitsos, E. I., et al. (2020). Tailoring the mechanical properties of metaluminous aluminosilicate glasses by phosphate incorporation. Front. Mat. 7 (115). doi:10.3389/fmats.2020.00115

CrossRef Full Text | Google Scholar

Green, D. J., Tandon, R., and Sglavo, V. M. (1999). Crack arrest and multiple cracking in glass through the use of designed residual stress profiles. Science 283 (5406), 1295–1297. doi:10.1126/science.283.5406.1295

PubMed Abstract | CrossRef Full Text | Google Scholar

Gross, T. M. (2019). “Chemical strengthening of glass,” in Springer handbook of glass. Editors J. D. Musgraves, J. Hu, and L. Calvez (Cham: Springer International Publishing), 273–296. doi:10.1007/978-3-319-93728-1_8

CrossRef Full Text | Google Scholar

Gross, T. M. (2012). Deformation and cracking behavior of glasses indented with diamond tips of various sharpness. J. Non-Cryst. Solids 358 (24), 3445–3452. doi:10.1016/j.jnoncrysol.2012.01.052

CrossRef Full Text | Google Scholar

Grund Bäck, L., Ali, S., Karlsson, S., Möncke, D., Kamitsos, E. I., and Jonson, B. (2019). Mixed alkali/alkaline earth-silicate glasses: Physical properties and structure by vibrational spectroscopy. Int. J. Appl. Glass Sci. 10 (3), 349–362. doi:10.1111/ijag.13101

CrossRef Full Text | Google Scholar

Güzel, A. S., Sarıgüzel, M., Özdemir Yanık, M. C., Günay, E., Usta, M., and Öztürk, Y. (2019). Enhancing mechanical endurance of chemical-tempered thin soda-lime silicate float glass by ion exchange. J. Aust. Ceram. Soc. 56, 185–201. doi:10.1007/s41779-019-00375-x

CrossRef Full Text | Google Scholar

Güzel, A. S., Sarıgüzel, M., Özdemir Yanık, M. C., Günay, E., Usta, M., and Öztürk, Y. (2020). Enhancing mechanical endurance of chemical-tempered thin soda-lime silicate float glass by ion exchange. J. Aust. Ceram. Soc. 56 (1), 185–201. doi:10.1007/s41779-019-00375-x

CrossRef Full Text | Google Scholar

Hand, R. J., and Tadjiev, D. R. (2010). Mechanical properties of silicate glasses as a function of composition. J. Non-Cryst. Solids 356 (44-49), 2417–2423. doi:10.1016/j.jnoncrysol.2010.05.007

CrossRef Full Text | Google Scholar

Hödemann, S., Valdmann, A., Anton, J., and Murata, T. (2016). Gradient scattered light method for non-destructive stress profile determination in chemically strengthened glass. J. Mat. Sci. 51, 5962–5978. doi:10.1007/s10853-016-9897-4

CrossRef Full Text | Google Scholar

Hödemann, S., Valdmann, A., Paemurru, M., Anton, J., Kiisk, V., Tkaczyk, E., et al. (2020). Measurement of stress build-up of ion exchange strengthened lithium aluminosilicate glass. J. Am. Ceram. Soc. 103 (4), 2407–2420. doi:10.1111/jace.16913

PubMed Abstract | CrossRef Full Text | Google Scholar

Iftekhar, S., Pahari, B., Okhotnikov, K., Jaworski, A., Stevensson, B., Grins, J., et al. (2012). Properties and structures of RE₂O₃-Al₂O₃-SiO₂ (RE = Y, Lu) glasses probed by molecular dynamics simulations and solid-state NMR: The roles of aluminum and rare-earth ions for dictating the microhardness. J. Phys. Chem. C 116 (34), 18394–18406. doi:10.1021/jp302672b

CrossRef Full Text | Google Scholar

Jacoby, M. (2018). Chemically strengthened glass’s next act: Car windshields. C&EN Glob. Enterp. 96 (3), 16–17. doi:10.1021/cen-09603-scitech1

CrossRef Full Text | Google Scholar

Jang, J.-i. (2009). Estimation of residual stress by instrumented indentation: A review. J. Ceram. Process. Res. 10 (3), 391–400. doi:10.36410/jcpr.2009.10.3.391

CrossRef Full Text | Google Scholar

Januchta, K., Youngman, R. E., Goel, A., Bauchy, M., Logunov, S. L., Rzoska, S. J., et al. (2017). Discovery of ultra-crack-resistant oxide glasses with adaptive networks. Chem. Mat. 29 (14), 5865–5876. doi:10.1021/acs.chemmater.7b00921

CrossRef Full Text | Google Scholar

Kang, M., Li, W., Leong, A. F. T., Guan, M., Fezzaa, K., Harris, J. T., et al. (2020). Crack nucleation and growth during dynamic indentation of chemically-strengthened glass. Extreme Mech. Lett. 38, 100754. doi:10.1016/j.eml.2020.100754

CrossRef Full Text | Google Scholar

Karlsson, S., Jonson, B., and Stålhandske, C. (2010). The technology of chemical glass strengthening - a review. Glass Technol. - Eur. J. Glass Sci. Technol. Part A 51 (2), 41–54.

Google Scholar

Karlsson, S. (2021). Viscosity of alumina doped soda lime silicate glasses – observation of anomaly in the linear increase as Al₂O₃ replaces SiO₂. J. Non-Cryst. Solids 573, 121149. doi:10.1016/j.jnoncrysol.2021.121149

CrossRef Full Text | Google Scholar

Karlsson, S. (2022). Compositional effects on indentation mechanical properties of chemically strengthened TiO₂-doped soda lime silicate glasses. Materials 15 (2), 577. doi:10.3390/ma15020577

PubMed Abstract | CrossRef Full Text | Google Scholar

Karlsson, S., Mathew, R., Ali, S., Paemurru, M., Stevensson, B., Anton, J., et al. (2022). Dataset: Mechanical, thermal, and structural investigations of chemically strengthened Na2O–CaO–Al2O3–SiO2 glasses. doi:10.5878/w6c3-hk86

CrossRef Full Text | Google Scholar

Karlsson, S., and Wondraczek, L. (2021). “Strengthening of oxide glasses,” in Encyclopedia for glass science, technology, history and culture. Editor P. Richet (Hoboken, NJ: John Wiley & Sons). doi:10.1002/9781118801017.ch3.12

CrossRef Full Text | Google Scholar

Kato, Y., Yamazaki, H., Yoshida, S., and Matsuoka, J. (2010). Effect of densification on crack initiation under Vickers indentation test. J. Non-Cryst. Solids 356 (35-36), 1768–1773. doi:10.1016/j.jnoncrysol.2010.07.015

CrossRef Full Text | Google Scholar

Kese, K., Li, Z., and Bergman, B. (2004). Influence of residual stress on elastic modulus and hardness of soda-lime glass measured by nanoindentation. J. Mat. Res. 19 (10), 3109–3119. doi:10.1557/JMR.2004.0404

CrossRef Full Text | Google Scholar

Kjeldsen, J., Smedskjaer, M. M., Mauro, J. C., and Yue, Y. (2014). On the origin of the mixed alkali effect on indentation in silicate glasses. J. Non-Cryst. Solids 406 (0), 22–26. doi:10.1016/j.jnoncrysol.2014.09.036

CrossRef Full Text | Google Scholar

LaCourse, W. C. (2018). “Design of SLS compositions for accelerated chemical strengthening,” in 78th conference on glass problems, 179–189. doi:10.1002/9781119519713.ch16

CrossRef Full Text | Google Scholar

Le Losq, C., Neuville, D. R., Florian, P., Henderson, G. S., and Massiot, D. (2014). The role of Al³⁺ on rheology and structural changes in sodium silicate and aluminosilicate glasses and melts. Geochim. Cosmochim. Acta 126, 495–517. doi:10.1016/j.gca.2013.11.010

CrossRef Full Text | Google Scholar

Lee, S. K. (2010). Effect of pressure on structure of oxide glasses at high pressure: Insights from solid-state NMR of quadrupolar nuclides. Solid State Nucl. Magn. Reson. 38 (2), 45–57. doi:10.1016/j.ssnmr.2010.10.002

PubMed Abstract | CrossRef Full Text | Google Scholar

Lee, S. K., and Sung, S. (2008). The effect of network-modifying cations on the structure and disorder in peralkaline Ca–Na aluminosilicate glasses: O-17 3QMAS NMR study. Chem. Geol. 256 (3–4), 326–333. doi:10.1016/j.chemgeo.2008.07.019

CrossRef Full Text | Google Scholar

Levitt, M. H., and Edén, M. (1998). Numerical simulation of periodic nuclear magnetic resonance problems: Fast calculation of carousel averages. Mol. Phys. 95 (5), 879–890. doi:10.1080/002689798166503

CrossRef Full Text | Google Scholar

Luo, J., Lezzi, P. J., Vargheese, K. D., Tandia, A., Harris, J. T., Gross, T. M., et al. (2016a). Competing indentation deformation mechanisms in glass using different strengthening methods. Front. Mat. 3 (52). doi:10.3389/fmats.2016.00052

CrossRef Full Text | Google Scholar

Luo, J., Vargheese, K. D., Tandia, A., Harris, J. T., and Mauro, J. C. (2016b). Structural origin of intrinsic ductility in binary aluminosilicate glasses. J. Non-Cryst. Solids 452, 297–306. doi:10.1016/j.jnoncrysol.2016.09.010

CrossRef Full Text | Google Scholar

Macrelli, G., Varshneya, A. K., and Mauro, J. C. (2019). Simulation of glass network evolution during chemical strengthening: Resolution of the subsurface compression maximum anomaly. J. Non-Cryst. Solids 522, 119457. doi:10.1016/j.jnoncrysol.2019.05.033

CrossRef Full Text | Google Scholar

Macrelli, G., Varshneya, A. K., and Mauro, J. C. (2020). Ultra-thin glass as a substrate for flexible photonics. Opt. Mater. 106, 109994. doi:10.1016/j.optmat.2020.109994

CrossRef Full Text | Google Scholar

Martin, B., Wondraczek, L., Deubener, J., and Yue, Y. (2005). Mechanically induced excess enthalpy in inorganic glasses. Appl. Phys. Lett. 86 (12), 121917. doi:10.1063/1.1895483

CrossRef Full Text | Google Scholar

Morandotti, A., and Zuccato, D. (2018). How to improve mechanical resistance of glass Syringes and Cartridges. PDA J. GMP Validation Jpn. 20 (1), 5–10. doi:10.11347/pda.20.5

CrossRef Full Text | Google Scholar

Morozumi, H., Nakano, H., Yoshida, S., and Matsuoka, J. (2015). Crack initiation tendency of chemically strengthened glasses. Int. J. Appl. Glass Sci. 6 (1), 64–71. doi:10.1111/ijag.12089

CrossRef Full Text | Google Scholar

Morris, D. J., Myers, S. B., and Cook, R. F. (2004). Indentation crack initiation in ion-exchanged aluminosilicate glass. J. Mater. Sci. 39 (7), 2399–2410. doi:10.1023/b:jmsc.0000020002.06117.46

CrossRef Full Text | Google Scholar

Musil, J., Kunc, F., Zeman, H., and Poláková, H. (2002). Relationships between hardness, Young's modulus and elastic recovery in hard nanocomposite coatings. Surf. Coatings Technol. 154 (2), 304–313. doi:10.1016/S0257-8972(01)01714-5

CrossRef Full Text | Google Scholar

Mysen, B. O., Virgo, D., and Kushiro, I. (1981). The structural role of aluminum in silicate melts—A Raman spectroscopic study at 1 atmosphere. Am. Mineralogist 66 (7-8), 678–701.

Google Scholar

Mysen, B. (2021). Structure of aluminosilicate melts. ISIJ Int. 61 (12), 2866–2881. doi:10.2355/isijinternational.ISIJINT-2021-100

CrossRef Full Text | Google Scholar

Oliver, W. C., and Pharr, G. M. (1992). An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J. Mat. Res. 7 (06), 1564–1583. doi:10.1557/JMR.1992.1564

CrossRef Full Text | Google Scholar

Pönitzsch, A., Nofz, M., Wondraczek, L., and Deubener, J. (2016). Bulk elastic properties, hardness and fatigue of calcium aluminosilicate glasses in the intermediate-silica range. J. Non-Cryst. Solids 434, 1–12. doi:10.1016/j.jnoncrysol.2015.12.002

CrossRef Full Text | Google Scholar

Ragoen, C., Cormier, L., Bidegaray, A. I., Vives, S., Henneman, F., Trcera, N., et al. (2018). A XANES investigation of the network-modifier cations environment before and after the Na⁺/K⁺ ion-exchange in silicate glasses. J. Non-Cryst. Solids 479, 97–104. doi:10.1016/j.jnoncrysol.2017.10.021

CrossRef Full Text | Google Scholar

Ragoen, C., Marple, M. A. T., Sen, S., Lambricht, T., and Godet, S. (2017a). Structural modifications induced by Na⁺/K⁺ ion exchange in silicate glasses: A multinuclear NMR spectroscopic study. J. Non-Cryst. Solids 474, 9–15. doi:10.1016/j.jnoncrysol.2017.08.006

CrossRef Full Text | Google Scholar

Ragoen, C., Sen, S., Lambricht, T., and Godet, S. (2017b). Effect of Al₂O₃ content on the mechanical and interdiffusional properties of ion-exchanged Na-aluminosilicate glasses. J. Non-Cryst. Solids 458, 129–136. doi:10.1016/j.jnoncrysol.2016.12.019

CrossRef Full Text | Google Scholar

Ray, N. H. (1974). Composition—Property relationships in inorganic oxide glasses. J. Non-Cryst. Solids 15 (3), 423–434. doi:10.1016/0022-3093(74)90148-3

CrossRef Full Text | Google Scholar

Ren, M., Cheng, J. Y., Jaccani, S. P., Kapoor, S., Youngman, R. E., Huang, L., et al. (2019). Composition – structure – property relationships in alkali aluminosilicate glasses: A combined experimental – computational approach towards designing functional glasses. J. Non-Cryst. Solids 505, 144–153. doi:10.1016/j.jnoncrysol.2018.10.053

CrossRef Full Text | Google Scholar

Rouxel, T. (2015). Driving force for indentation cracking in glass: Composition, pressure and temperature dependence. Phil. Trans. R. Soc. A 373, 20140140. (2038). doi:10.1098/rsta.2014.0140

PubMed Abstract | CrossRef Full Text | Google Scholar

Rouxel, T. (2007). Elastic properties and short-to medium-range order in glasses. J. Am. Ceram. Soc. 90 (10), 3019–3039. doi:10.1111/j.1551-2916.2007.01945.x

CrossRef Full Text | Google Scholar

Rouxel, T., Jang, J.-i., and Ramamurty, U. (2021). Indentation of glasses. Prog. Mater. Sci. 121, 100834. doi:10.1016/j.pmatsci.2021.100834

CrossRef Full Text | Google Scholar

Rouxel, T., Ji, H., Hammouda, T., and Moreac, A. (2008). Poisson's ratio and the densification of glass under high pressure. Phys. Rev. Lett. 100 (22), 225501–225504. doi:10.1103/Physrevlett.100.225501

PubMed Abstract | CrossRef Full Text | Google Scholar

Ryou, S.-Y., Lee, C.-S., Cho, I.-S., and Amanov, A. (2020). Measurement of dynamic elastic modulus and Poisson’s ratio of chemically strengthened glass. Materials 13 (24), 5644. doi:10.3390/ma13245644

CrossRef Full Text | Google Scholar

Shannon, R. (1976). Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Cryst. Sect. A 32 (5), 751–767. doi:10.1107/S0567739476001551

CrossRef Full Text | Google Scholar

Sheikh, M. Z., Atif, M., Raza, M. A., Suo, T., Li, Y., Zhou, F., et al. (2020). Damage propagation and dynamic material properties of aluminosilicate glass. J. Non-Cryst. Solids 547, 120313. doi:10.1016/j.jnoncrysol.2020.120313

CrossRef Full Text | Google Scholar

Shelby, J. E. (2005). Introduction to glass science and technology. Cambridge, United Kingdom: Royal Society of Chemistry. doi:10.1039/9781847551160

CrossRef Full Text | Google Scholar

Silverman, W. (1939). Effect of alumina on devitrification of soda‐lime‐silica glasses. J. Am. Ceram. Soc. 22, 378–384. doi:10.1111/j.1151-2916.1939.tb19484.x

CrossRef Full Text | Google Scholar

Smedskjaer, M. M., Mauro, J. C., Kjeldsen, J., and Yue, Y. (2013). Microscopic origins of compositional trends in aluminosilicate glass properties. J. Am. Ceram. Soc. 96 (5), 1436–1443. doi:10.1111/jace.12298

CrossRef Full Text | Google Scholar

Stevensson, B., and Edén, M. (2006). Efficient orientational averaging by the extension of Lebedev grids via regularized octahedral symmetry expansion. J. Magnetic Reson. 181 (1), 162–176. doi:10.1016/j.jmr.2006.04.008

PubMed Abstract | CrossRef Full Text | Google Scholar

Sun, H., and Dugnani, R. (2020). A study on ion-exchanged, soda-lime glass’s residual stress relationship with K⁺/Na⁺ concentration. Int. J. Appl. Glass Sci. 11 (1), 134–146. doi:10.1111/ijag.14077

CrossRef Full Text | Google Scholar

Sun, H., Hödemann, S., and Dugnani, R. (2019). Characterization of shallow stress-profiles in chemically strengthened soda-lime glass. J. Non-Cryst. Solids 510, 130–142. doi:10.1016/j.jnoncrysol.2019.01.023

CrossRef Full Text | Google Scholar

Sundberg, P., Grund Bäck, L., Orman, R., Booth, J., and Karlsson, S. (2019). Simultaneous chemical vapor deposition and thermal strengthening of glass. Thin Solid Films 669, 487–493. doi:10.1016/j.tsf.2018.11.028

CrossRef Full Text | Google Scholar

Svenson, M. N., Bechgaard, T. K., Fuglsang, S. D., Pedersen, R. H., Tjell, A. Ø., Østergaard, M. B., et al. (2014). Composition-structure-property relations of compressed borosilicate glasses. Phys. Rev. Appl. 2 (2), 024006. doi:10.1103/PhysRevApplied.2.024006

CrossRef Full Text | Google Scholar

Svenson, M. N., Mauro, J. C., Rzoska, S. J., Bockowski, M., and Smedskjaer, M. M. (2017). Accessing forbidden glass regimes through high-pressure sub-t_g annealing. Sci. Rep. 7, 46631. doi:10.1038/srep46631

PubMed Abstract | CrossRef Full Text | Google Scholar

Svenson, M. N., Thirion, L. M., Youngman, R. E., Mauro, J. C., Bauchy, M., Rzoska, S. J., et al. (2016). Effects of thermal and pressure histories on the chemical strengthening of sodium aluminosilicate glass. Front. Mat. 3, 14. doi:10.3389/fmats.2016.00014

CrossRef Full Text | Google Scholar

Takahashi, S., Neuville, D. R., and Takebe, H. (2015). Thermal properties, density and structure of percalcic and peraluminus CaO–Al₂O₃–SiO₂ glasses. J. Non-Cryst. Solids 411 (0), 5–12. doi:10.1016/j.jnoncrysol.2014.12.019

CrossRef Full Text | Google Scholar

Talimian, A., and Sglavo, V. M. (2017). Ion-exchange strengthening of borosilicate glass: Influence of salt impurities and treatment temperature. J. Non-Cryst. Solids 456, 12–21. doi:10.1016/j.jnoncrysol.2016.10.032

CrossRef Full Text | Google Scholar

Tandia, A., Vargheese, K. D., Mauro, J. C., and Varshneya, A. K. (2012). Atomistic understanding of the network dilation anomaly in ion-exchanged glass. J. Non-Cryst. Solids 358 (2), 316–320. doi:10.1016/j.jnoncrysol.2011.09.034

CrossRef Full Text | Google Scholar

Toplis, M. J., Dingwell, D. B., and Lenci, T. (1997). Peraluminous viscosity maxima in Na₂O-Al₂O₃-SiO₂ liquids: The role of triclusters in tectosilicate melts. Geochim. Cosmochim. Acta 61 (13), 2605–2612. doi:10.1016/S0016-7037(97)00126-9

CrossRef Full Text | Google Scholar

Varshneya, A. K., and Bihuniak, P. P. (2018). Cover screens for personal electronic devices: Strengthened glass or sapphire. Am. Ceram. Soc. Bull. 96, 21–25.

Google Scholar

Varshneya, A. K., and Kreski, P. K. (2012). “The chemistry of chemical strengthening of glass,” in Processing, properties, and applications of glass and optical materials. Editors A. K. Varshneya, H. A. Schaeffer, K. A. Richardson, M. Wightman, and L. D. Pye (Hoboken, New Jersey: John Wiley & Sons), 107–114. doi:10.1002/9781118472590.ch12

CrossRef Full Text | Google Scholar

Varshneya, A. K., Olson, G. A., Kreski, P. K., and Gupta, P. K. (2015). Buildup and relaxation of stress in chemically strengthened glass. J. Non-Cryst. Solids 427, 91–97. doi:10.1016/j.jnoncrysol.2015.07.037

CrossRef Full Text | Google Scholar

Varshneya, A. K., Macrelli, G., Yoshida, S., Kim, S. H., Ogrinc, A. L., and Mauro, J. C. (2022). Indentation and abrasion in glass products: Lessons learned and yet to be learned. Int. J. Appl. Glass Sci. 13, 308–337. doi:10.1111/ijag.16549

CrossRef Full Text | Google Scholar

Varshneya, A. K. (2010a). Chemical strengthening of glass: Lessons learned and yet to Be learned. Int. J. Appl. Glass Sci. 1 (2), 131–142. doi:10.1111/j.2041-1294.2010.00010.x

CrossRef Full Text | Google Scholar

Varshneya, A. K. (2010b). The physics of chemical strengthening of glass: Room for a new view. J. Non-Cryst. Solids 356 (44-49), 2289–2294. doi:10.1016/j.jnoncrysol.2010.05.010

CrossRef Full Text | Google Scholar

Varshneya, A. K. (2016). Mechanical model to simulate buildup and relaxation of stress during glass chemical strengthening. J. Non-Cryst. Solids 433, 28–30. doi:10.1016/j.jnoncrysol.2015.11.006

CrossRef Full Text | Google Scholar

Wallenberger, F. T., and Bingham, P. A. (2010). Fiberglass and glass technology: Energy-friendly compositions. New York: Springer. doi:10.1007/978-1-4419-0736-3

CrossRef Full Text | Google Scholar

Wang, M., Smedskjaer, M. M., Mauro, J. C., Sant, G., and Bauchy, M. (2017). Topological origin of the network dilation anomaly in ion-exchanged glasses. Phys. Rev. Appl. 8 (5), 054040. doi:10.1103/PhysRevApplied.8.054040

CrossRef Full Text | Google Scholar

Webb, S. L., Banaszak, M., Köhler, U., Rausch, S., and Raschke, G. (2007). The viscosity of Na₂O-CaO-Al₂O₃-SiO₂ melts. Eur. J. Mineral. 19 (5), 681–692. doi:10.1127/0935-1221/2007/0019-1765

CrossRef Full Text | Google Scholar

Weigel, C., Le Losq, C., Vialla, R., Dupas, C., Clément, S., Neuville, D. R., et al. (2016). Elastic moduli of XAlSiO₄ aluminosilicate glasses: Effects of charge-balancing cations. J. Non-Cryst. Solids 447, 267–272. doi:10.1016/j.jnoncrysol.2016.06.023

CrossRef Full Text | Google Scholar

Wondraczek, L., Mauro, J. C., Eckert, J., Kühn, U., Horbach, J., Deubener, J., et al. (2011). Towards ultrastrong glasses. Adv. Mat. 23 (39), 4578–4586. doi:10.1002/adma.201102795

PubMed Abstract | CrossRef Full Text | Google Scholar

Xiaoyu, L., Liangbao, J., Iman, M., John, P. D., Lei, L., and Yue, Y. (2018). New insights into nanoindentation crack initiation in ion‐exchanged sodium aluminosilicate glass. J. Am. Ceram. Soc. 101 (7), 2930–2940. doi:10.1111/jace.15465

CrossRef Full Text | Google Scholar

Xue, X., and Stebbins, J. F. (1993). ²³Na NMR chemical shifts and local Na coordination environments in silicate crystals, melts and glasses. Phys. Chem. Min. 20 (5), 297–307. doi:10.1007/BF00215100

CrossRef Full Text | Google Scholar

Yetna N'Jock, M., Roudet, F., Idriss, M., Bartier, O., and Chicot, D. (2016). Work-of-indentation coupled to contact stiffness for calculating elastic modulus by instrumented indentation. Mech. Mater. 94, 170–179. doi:10.1016/j.mechmat.2015.12.003

CrossRef Full Text | Google Scholar

Yoshida, S., Hidaka, A., and Matsuoka, J. (2004). Crack initiation behavior of sodium aluminosilicate glasses. J. Non-Cryst. Solids 344 (1-2), 37–43. doi:10.1016/J.Jnoncrysol.2004.07.019

CrossRef Full Text | Google Scholar

Yoshida, S. (2019). Indentation deformation and cracking in oxide glass –toward understanding of crack nucleation. J. Non-Cryst. Solids X 1, 100009. doi:10.1016/j.nocx.2019.100009

CrossRef Full Text | Google Scholar

Yu, Y., Stevensson, B., and Edén, M. (2020). A unified ²³Na NMR chemical shift correlation with structural parameters in multicomponent silicate-based glasses. J. Am. Ceram. Soc. 103 (2), 762–767. doi:10.1111/jace.16852

CrossRef Full Text | Google Scholar

Yue, Y. Z., Christiansen, J. d., and Jensen, S. L. (2002). Determination of the fictive temperature for a hyperquenched glass. Chem. Phys. Lett. 357 (1), 20–24. doi:10.1016/S0009-2614(02)00434-7

CrossRef Full Text | Google Scholar

Zaccaria, M., Dubru, M., Lucca, N., and Šikyňová, A. (2021). Chemically strengthened glass for architectural applications. Ce. Pap. 4 (6), 135–144. doi:10.1002/cepa.1626

CrossRef Full Text | Google Scholar

Zanotto, E. D., and Mauro, J. C. (2017). The glassy state of matter: Its definition and ultimate fate. J. Non-Cryst. Solids 471, 490–495. doi:10.1016/j.jnoncrysol.2017.05.019

CrossRef Full Text | Google Scholar

Zhang, Z., Soga, N., and Hirao, K. (1995). Indentation deformation and fracture of densified silicate glass. J. Mater. Sci. 30 (24), 6359–6362. doi:10.1007/BF00369689

CrossRef Full Text | Google Scholar

Zheng, Q., Zhang, Y., Montazerian, M., Gulbiten, O., Mauro, J. C., Zanotto, E. D., et al. (2019). Understanding glass through differential scanning calorimetry. Chem. Rev. 119 (13), 7848–7939. doi:10.1021/acs.chemrev.8b00510

PubMed Abstract | CrossRef Full Text | Google Scholar

Zheng, Q., Zheng, J., Solvang, M., Yue, Y., and Mauro, J. C. (2020). Determining the liquidus viscosity of glass-forming liquids through differential scanning calorimetry. J. Am. Ceram. Soc. 103 (11), 6070–6074. doi:10.1111/jace.17363

CrossRef Full Text | Google Scholar