Properties of Ca–(Y)–Si–Al–O–N–F Glasses: Independent and Additive Effects of Fluorine and Nitrogen A. R. Garc  ıa-Bell  es, ‡ M. Monz o, ‡ A. Barba, ‡ C. Clausell, ‡,† M. J. Pomeroy, § A. R. Hanifi, § and S. Hampshire § ‡ Instituto Universitario de Tecnologı´a Cera´mica (IUTC), Universitat Jaume I, Castell on 12071, Spain § Materials and Surface Science Institute, University of Limerick, Limerick, Ireland Thirty glasses of composition (in equivalent percent) 20-xCa: xY:50Si:30Al:(100-y-z)O:yN:zF, with x = 0, 10; y = 0, 10, 20, and z = 0, 1, 3, 5, 7 were prepared by melting and casting. All glasses were X-ray amorphous. Glass molar volumes (MV) decreased with nitrogen substitution for oxygen for all fluorine contents and, correspondingly, glass fractional compactness increased. Fluorine substitution of oxygen had virtually no effect on molar volume or fractional glass compactness for the three nitrogen contents tested. Young’s modulus and microh- ardness were virtually unaffected by fluorine substitution for oxygen while nitrogen substitution for oxygen caused increases in these two properties. Glass-transition temperature and dila- tometric-softening point values all decreased with increasing fluorine substitution levels, while increasing nitrogen substitu- tion caused values for these thermal properties to increase. Correspondingly, the thermal expansion coefficient increased with fluorine and decreased with nitrogen substitution levels. Using property value differences between glasses containing fluorine and the corresponding glass containing 0 eq.% F enabled 24 data points to be used to determine the effect of fluorine on T g,dil and T DS . The trends were linear with a gradi- ent for both properties of the order of 22°C (eq.% F) -1 . For the nitrogen effect, 20 data points were analyzed for trend effects. As expected from earlier work, all trends had good lin- earity. Gradients were for T g,dil and T DS +2.5°C (eq.% N) -1 , which are fairly similar to previous results in oxynitride sys- tems. All of the data collected and its analysis clearly shows that the substitution effects of fluorine for oxygen and nitrogen for oxygen are independent and additive with the fluorine sub- stitution. The property trends of the glasses are discussed in terms of their implications for glass structure. I. Introduction F LUORINE-CONTAINING glasses are used for a wide variety of purposes, among them bioglasses and bioglass ceram- ics, where fluoride release stimulates hydroxyapatite forma- tion, 1 which bonds to human bone due to similar phase structure. 2–4 Fluorine is also introduced into ionomer glasses which are used for glass polyalkenoate dental cements, where fluorine atoms are added to lower the refractive index of the glass as well as to enable fluoride ion release from the set cement 5 to prevent secondary caries. 6 Fluorine ions in human saliva and plasma also play an important role in develop- ment of hard tissues in the body. 7 It is well-known that fluorine facilitates melting of glass at lower temperatures and acts as a powerful network disrupter, 8,9 creating a marked reduction in the glass- transition temperature (T g ), viscosity, and refractive index as the fluorine content of the glasses increases. These effects are explained on the basis of replacing bridging oxygens by non- bridging, terminating fluorines, thereby reducing the network connectivity and facilitating network mobility at lower tem- peratures. Fluorine also aids crystallization and increases the potential for phase separation. 10–12 The electrical conductiv- ity and the thermal expansion coefficient of oxyfluoride glasses are raised as the fluorine content increases. 13 It has been suggested that fluorine can exist as fluorite (CaF 2 ) clus- ters in calcium-modified glasses, 14 bound to silicon as Si–F or bound to Al as Al–F species. 9,15,16 Si–F bonding can lead to formation of volatile SiF 4 , which must be avoided as it is associated with both fluorine loss and an environmental problem as the SiF 4 produced will hydrolyze in the presence of water to hydrofluoric acid and silica. Some studies have been carried out on the structural role of fluorine in glasses. 9,10,15,17 Despite evidence of Si–F bonds in silicate or alumino–silicate glasses, it has been suggested 14 that fluorine loss as SiF 4 could be suppressed completely from fluoro– alumino–silicate glasses by appropriate choice of composition and the incorporation of sufficient basic network modifying oxide. The ratios between Si:F and Al:F are important in determining the existence of Si–F bonds, with only small quan- tities forming when Al:F  1 (i.e., when there are enough Al atoms to satisfy all fluorine atoms). Structural studies of fluo- rine-containing alumino–silicate glasses have shown the pres- ence of F–Ca(n) and Al–F–Ca(n) species in these glasses. 10,17 On the other hand, numerous investigations have been carried out on glass formation and properties in a wide range of M–Si– O–N and M–Si–Al–O–N glasses, where “M” is typically an alkali, alkaline-earth or rare-earth metal and a comprehensive review is given by Becher et al. 18 In the initial studies, 19–22 corre- lations between the amount of silicon nitride dissolved into oxy- nitride glasses and changes in their physical properties were reported. Glass-transition temperature, microhardness, and rela- tive fracture toughness all increased with increasing nitrogen con- tent while the thermal expansion coefficient decreased. IR spectroscopic analyses 21 indicated that the incorporated nitrogen became chemically bonded to silicon in the glass network and, by substitution for oxygen, produced a more tightly and highly linked structure. However, the cation ratios for these glasses were variable and it was not possible to say unequivocally that the improvements in properties observed were solely due to the increased nitrogen concentration in the glass. In contrast, Drew 23,24 and Hampshire 25–27 carried out extensive studies on glasses in M–Si–O–N and M–Si–Al–O–N (where M = Ca, Mg, This article was published online on 15 March 2013. An error was subsequently identi- fied. This notice is included in the online and print versions to indicate that both have been corrected 26 March 2013. H. J. Kleebe—contributing editor Manuscript No. 32121. Received October 2, 2012; approved January 27, 2013. † Author to whom correspondence should be addressed. e-mail: carola.clausell@ uji.es 1131 J. Am. Ceram. Soc., 96 [4] 1131–1137 (2013) DOI: 10.1111/jace.12249 © 2013 The American Ceramic Society J ournal