We thank the Galuskins for their detailed study of the explosion breccias and the nitrides included in corundum aggregates from Mt Carmel; space considerations have limited our previous publication of such detailed data on this interesting aspect of these important samples. Their images of other samples of the ‘Carmel Sapphire’ are a useful supplement to those we have published elsewhere.
However, we deem it necessary to correct some unfortunate mistakes in the presentation. These do not affect the descriptions of the images but can improve the usefulness of the article. Material in quotation marks are from the manuscript in Mineralogical Magazine, vol. 87, https://doi.org/10.1180/mgm.2023.25 by Galuskin and Galuskina (Reference Galuskin and Galuskina2023).
1. The authors do not identify the location of their sample (there is no “Carmel locality at the Kishon River”), the methods by which it was obtained, nor the other minerals found in the heavy concentrate. This is unfortunate because it would have improved the scientific usefulness of the paper.
2. “Typically the Carmel Sapphire is in small fragments of breccia with white cement”. This is incorrect; the vast bulk of Carmel sapphire grains investigated by us have no coating and are not associated with the breccias.
3. “skeletal osbornite occurring as inclusions in corundum”. The illustrated nitrides are not ‘skeletal crystals’ – they may appear to be so because they are controlled by crystallographically controlled void spaces in skeletal corundum (Griffin et al., Reference Griffin, Gain, Adams, Huang, Saunders, Toledo, Pearson and O'Reilly2016; Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022; Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). These spectacular structures are only one form of TiN in the Carmel sapphire; it also occurs in large inclusions coexisting with Fe–Ti silicides and TiB2, and as regular crystals that appear to have crystallised directly from Fe–Ti silicide melts (Griffin et al., Reference Griffin, Gain, Camara, Bindi, Shaw, Alard, Saunders, Huang, Toledo and O'Reilly2020a, Reference Griffin, Gain, Saunders, Bindi, Alard, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2022).
4. “as a result of CH4+H2 fluid flowing through the magmatic melt”. This is not quite correct; we have proposed that such fluids interacted with melts, and may have flowed through melt-escape channels together with melts, to produce the skeletal crystals of the Carmel sapphire (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021).
5. “On only one occasion the amorphous carbon is recorded in the cement of the ‘white breccia’.” The authors confuse the white material that coats some grains with the explosion breccias illustrated by us (Xiong et al., Reference Xiong, Griffin, Huang, Gain, Toledo, Pearson and O'Reilly2017; Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b). The matrix of the explosion breccias, which may include the one large grain described by the authors, are dominated by amorphous carbon, as identified by petrographic, EBSD and TEM studies. We have noted micro-inclusions of corundum, but not bauxite, in the matrix of the explosion breccias. The authors provide no evidence that bauxite, or any other Al-hydroxide, is present in the matrices. The EDS spectra given for “böhmite-like material” are equally consistent with corundum. We suggest that the report of böhmite /bauxite represents analyses of spots consisting of amorphous carbon plus micro-inclusions of corundum; this would be consistent with the apparently higher C content of the EDS spectrum. If böhmite actually is present, it may equally reflect partial alteration of corundum fragments. Whether this fine-grained material could be considered as an ash component is a semantic question.
6. “Osbornite from Carmel sapphire grains…as a rule are confined to glass inclusions”. This is not correct. The osbornite ‘skeletal crystals’ typically extend well outside the melt pockets, and occur as isolated inclusions in the corundum (as seen in 3D-μCT as well as 2-D slices; Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022). See also point (3) above.
7. “Griffin et al. (Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b) underline the non-stable composition of osbornite”. This is simply not correct; we have done no such thing. The cited reference presents analytical data and shows that they are consistent with experimental studies on the Ti–O–N system. There is nothing “non-stable” there.
8. “Bigger corundum aggregates contain silica glass”; “usually squeezed out between corundum grains”. This is misleading; the glasses in the melt pockets are calcium–aluminium–silicate (CAS) glasses (typically with 40–45 wt.% SiO2) with moderate levels of Mg, Ti and S, as well as high levels of REE and Zr. In some parageneses with cumulate-like structures the melts are indeed concentrated along grain boundaries, whereas in the more skeletal corundum aggregates the melts occur both along grain boundaries and as isolated inclusions. 3μCT images (Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a) illustrate this very clearly.
9. “.. ‘white breccia’ … as a source of Carmel sapphire”. This is misleading; nearly all investigated grains of Carmel sapphire have no traces of the white coating, and there is no evidence that they have been brecciated.
10. Zr-bearing glass as debris from “special laboratory glassware”. As noted above, the internal melts in the Carmel sapphire are high in Zr and the Carmel sapphire contains many Zr-bearing phases. Some corundum-aggregate xenoliths have thick rims of a Ba-rich or K-rich glass full of tiny crystals of zircon or baddeleyite. The high Zr contents reflect the fact that most of the melts left in the corundum aggregates by the time of eruption were highly differentiated. The spectrum in fig. 7d in the author's manuscript is very similar to those of the glasses in melt pockets in one type of Carmel Sapphire. It is unfortunate that these spectra were not accompanied by quantitative chemical analyses, which would have aided comparison with published data.
11. “.. terrestrial examples [of krotite and grossite] are questionable and require verification.” The authors might read our papers (Griffin et al. Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2020a) on the coarse-grained xenoliths of hibonite+ grossite+V0, which require a very H2-rich atmosphere and show the crystallisation sequence:
corundum + Liq → (low-REE) hibonite → grossite + spinel ± krotite → Ca4Al6F2O12 + fluorite. The occurrence of grossite in meteorites cannot invalidate the existence of these samples; their disaggregation during eruption is responsible for inclusions of krotite, hibonite and grossite in the breccias. These clearly did not result from the hydration of hypothetical precursor minerals.
12. “Glass fragments with inclusions of mullite”. Mullite does appear in some melt inclusions in the corundum aggregates. This is expected because the compositions of some residual melts fall into the liquidus field of mullite in CAS space at low P, and the crystallisation of mullite might be expected to occur during eruption.
13. “There is no mechanism for the simultaneous removal of Fe and Si from natural melt systems”. This is not correct. Reduction of an FeO-bearing silicate melt to f O2 = IW-5 (as documented by the abundance of Ti3+-bearing minerals in Carmel sapphire) will lead to immiscibility of Fe–Ti–Si silicides and supersaturation in Al2O3, leading to corundum crystallisation. Comparison of the analysed silicide-melt compositions with experimental data shows that this occurred at normal magmatic temperatures (1200–1400°C; Weitzer et al., Reference Weitzer, Schuster, Naka, Stein and Palm2008; Griffin et al., Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022). It is also worth noting that the presence of hydrogen can lower the melting points of metals and alloys by several hundred degrees (Fukai, Reference Fukai2005). Thus the 2000°C temperature of the industrial process is irrelevant to the natural case, although the observation that the silicide melts do sink to the bottom is a nice illustration of the process that we have proposed. We would appreciate a reference to the industrial process, which was not provided in the paper.
14. “formation from the melt requires extremely rapid quenching”. We agree; this is consistent with the very rapid eruption rate of the host basalts (pyroclastics) of the Carmel sapphire, which carry mantle-derived peridotite xenoliths (O'Reilly et al., Reference O'Reilly, Griffin, Dosetto, Turner and Van Orman2010)
15. “Therefore all new minerals [from] Carmel sapphire should be discredited by the IMA–CNMNC”. This is surely misguided (see below) and in any case is not appropriate to a journal article.
Finally, we must point out that a similarity between synthetic products (laboratory or industrial) and natural materials provides keys to the interpretation of natural processes (indeed, it is the raison d'etre of experimental petrology). However, such similarity cannot be taken as evidence for anthropogenic origin of the natural material. Duplication of mantle minerals in experiments does not imply that the minerals cannot exist in nature. This is the fundamental fallacy of the persistent criticisms by Litasov et al. (Reference Litasov, Kagi and Bekker2019a,Reference Litasov, Bekker and Kagib,Reference Litasov, Kagi, Voropaev, Hirata, Ohfuji, Ishibashi, Makino, Bekker, Sevastyanov, Afanasiev and Pokhilenkoc) and Ballhaus et al. (Reference Ballhaus, Wirth, Fonseca, Blanchard, Pröll, Bragagni, Nagel, Schreiber, Dittrich and Thome2017, Reference Ballhaus, Helmy, Fonseca, Wirth, Schreiber and Jöns2021) of the Mt Carmel work and many studies of similar mineral associations from numerous ophiolites by many researchers worldwide. Unfortunately, the experimental petrologists involved in this criticism have not followed scientific method; that would have included taking into account the overwhelming geological and geochemical evidence that makes an anthropogenic origin impossible, at least in the case of Mt Carmel (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
This evidence includes:
• The alluvial deposits in question are undisturbed Plio–Pleistocene palaeo-placers developed at the base of terraces that lie ca 10 metres above the current drainage of the Kishon River and are overlain by 4–10 metres of undisturbed finer-grained sediments (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
• Extensive sampling of known Cretaceous volcanic centres, which lie 50–350 metres above the Kishon River, shows that the vents collectively contain all of the minerals found in the alluvial deposits (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b), including moissanite and “Carmel Sapphire”.
• Miocene sedimentary deposits in the Yizre'l valley include carbonate-cemented, quartz-rich beach placers containing corundum, coarse-grained moissanite and peridotite-derived garnet+cpx+spinel (Roup et al., Reference Roup, Kamanovitch, Baykov and Toledo2009). Anthropogenic industrial processes were not available at this time.
• The occurrence of the corundum-moissanite suite in the primary volcanic centres, in deeply buried Plio–Pleistocene placers, and in lithified Miocene beach placers demonstrates that the sample material was formed and deposited long before human occupation and makes arguments for industrial contamination unsustainable (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
• The detailed descriptions of parageneses (Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b, Reference Griffin, Toledo and O'Reilly2019b, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a,Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reillyb, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022 and references therein) should lay to rest claims that such material was produced by lightning strikes (Ballhaus et al., Reference Ballhaus, Wirth, Fonseca, Blanchard, Pröll, Bragagni, Nagel, Schreiber, Dittrich and Thome2017).
• Cathodoluminescence imagery and Ti-zoning of corundum crystals (“Carmel Sapphire”) record growth from a silicate melt in an evolving open system at very low oxygen fugacity (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). This is not consistent with any industrial process known to us.
• In the Mt Carmel material, moissanite occurs in grains much larger than any common industrial abrasive (euhedral crystals up to > 4mm; Huang et al., Reference Huang, Xiong, Gain, Griffin, Murphy, Siryaev, Li, Toledo, Tomshin and O'Reilly2020). Aggregates of “Carmel Sapphire” and associated hibonite-grossite-vanadium assemblages are up to 2.5 cm across (Griffin et al., Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a, Reference Griffin, Gain, Camara, Bindi, Shaw, Alard, Saunders, Huang, Toledo and O'Reilly2020a).
The Mt Carmel mineral assemblages show significant differences to the synthetic materials produced and/or described by those arguing for a contamination origin (Litasov et al., Reference Litasov, Kagi and Bekker2019a,Reference Litasov, Bekker and Kagib,Reference Litasov, Kagi, Voropaev, Hirata, Ohfuji, Ishibashi, Makino, Bekker, Sevastyanov, Afanasiev and Pokhilenkoc).
• Tistarite, Ti2O3, in Carmel Sapphire does not have the high Al and Mg contents usually present in the synthetic material (Griffin et al., Reference Griffin, Gain, Adams, Huang, Saunders, Toledo, Pearson and O'Reilly2016).
• Synthetic ‘carmeltazite’ (Litasov et al., Reference Litasov, Kagi and Bekker2019a) shows a wide range in Zr content, requiring a high proportion of Ti4+ for charge balance. EELS analysis and single-crystal XRD analysis of the type carmeltazite from Mt Carmel show that all Ti is present as Ti3+ (Griffin et al., Reference Griffin, Gain, Bindi, Toledo, Camara, Saunders and O'Reilly2018a).
• Synthetic TiB2 in fused abrasives (Litasov et al., Reference Litasov, Kagi and Bekker2019a) contains significant levels of Al and F, which are absent in the TiB2 from Mt Carmel (Griffin et al., Reference Griffin, Gain, Saunders, Bindi, Alard, Toledo and O'Reilly2020b).
• The only EMPA-detectable minor elements in Carmel Sapphire are Ti, Si and Mg, however it contains trace elements (Sc, V, Cr and Zr), most of which are correlated with Ti contents ranging up to 2.5 wt.%. (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). In contrast, synthetic corundum produced from fused abrasives (Litasov et al., Reference Litasov, Kagi and Bekker2019a) contains much less Ti and no Mg, however it has significant contents of Fe, Ni, Mn, Ca, K, Th and P, which are absent or present only at low trace-element levels in ‘Carmel Sapphire’.
• TiC in the fused abrasives analysed by Litasov et al. (Reference Litasov, Kagi and Bekker2019a) contains no Fe, whereas that in ‘Carmel sapphire’ typically contains Fe at ca 1 wt.% levels, reflecting equilibrium partitioning with coexisting Fe–Si melts (Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022).
• The most reduced of the Mt Carmel parageneses (corundum + hibonite + grossite + native vanadium) requires an atmosphere dominated by H2. To our knowledge, this does not occur in the production of alumina or Al2O3-based abrasives, because Al2O3 would be reduced to the metal, which is undesirable. However, it is a predictable result of the ascent and partial oxidation of COH fluids from a metal-saturated mantle (Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b, Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a).
• The oxygen-isotope composition of Carmel Sapphire ranges from δ18O = +7 to +13‰, with a mean around +10 (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). Natural corundum shows a range of δ18O from –27 to +23‰. In contrast, commercial synthetic-alumina products produced by a range of processes have δ18O = +23 to +32‰, reflecting the interaction of the material with air (δ18O = 23.9±0.3 ‰) during processing (Pack, Reference Pack, Bindeman and Pack2021). This difference is critical in distinguishing natural corundum derived from mantle/magma-related processes from a synthetic product.
• U-rich thorianite [(Th,U)O2] occurs with ziroite (ZrO2), mizraite-(Ce) [Ce(Al11Mg)O19] and Ti-sulfide in one melt inclusion in a Carmel Sapphire. EMPA Th–U–Pb age determination of the largest 1-μm thorianite crystal using a JEOL JXA-iHP200F Field Emission EPMA and the Probe for EPMA program yields an age of 13±2 Ma (Ma et al., pers com.), making it unlikely that the corundum host was produced by humans.
This evidence indicates that there is no scientific basis for the authors’ conclusion that “the contamination of geological samples with anthropogenic material has led to popularisation of biased views”. The responsibility for “popularisation of biased views” must lie solely with those who refuse to consider straightforward geological and geochemical evidence for the natural origin of the Mt Carmel samples.
We thank the Galuskins for their detailed study of the explosion breccias and the nitrides included in corundum aggregates from Mt Carmel; space considerations have limited our previous publication of such detailed data on this interesting aspect of these important samples. Their images of other samples of the ‘Carmel Sapphire’ are a useful supplement to those we have published elsewhere.
However, we deem it necessary to correct some unfortunate mistakes in the presentation. These do not affect the descriptions of the images but can improve the usefulness of the article. Material in quotation marks are from the manuscript in Mineralogical Magazine, vol. 87, https://doi.org/10.1180/mgm.2023.25 by Galuskin and Galuskina (Reference Galuskin and Galuskina2023).
1. The authors do not identify the location of their sample (there is no “Carmel locality at the Kishon River”), the methods by which it was obtained, nor the other minerals found in the heavy concentrate. This is unfortunate because it would have improved the scientific usefulness of the paper.
2. “Typically the Carmel Sapphire is in small fragments of breccia with white cement”. This is incorrect; the vast bulk of Carmel sapphire grains investigated by us have no coating and are not associated with the breccias.
3. “skeletal osbornite occurring as inclusions in corundum”. The illustrated nitrides are not ‘skeletal crystals’ – they may appear to be so because they are controlled by crystallographically controlled void spaces in skeletal corundum (Griffin et al., Reference Griffin, Gain, Adams, Huang, Saunders, Toledo, Pearson and O'Reilly2016; Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022; Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). These spectacular structures are only one form of TiN in the Carmel sapphire; it also occurs in large inclusions coexisting with Fe–Ti silicides and TiB2, and as regular crystals that appear to have crystallised directly from Fe–Ti silicide melts (Griffin et al., Reference Griffin, Gain, Camara, Bindi, Shaw, Alard, Saunders, Huang, Toledo and O'Reilly2020a, Reference Griffin, Gain, Saunders, Bindi, Alard, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2022).
4. “as a result of CH4+H2 fluid flowing through the magmatic melt”. This is not quite correct; we have proposed that such fluids interacted with melts, and may have flowed through melt-escape channels together with melts, to produce the skeletal crystals of the Carmel sapphire (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021).
5. “On only one occasion the amorphous carbon is recorded in the cement of the ‘white breccia’.” The authors confuse the white material that coats some grains with the explosion breccias illustrated by us (Xiong et al., Reference Xiong, Griffin, Huang, Gain, Toledo, Pearson and O'Reilly2017; Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b). The matrix of the explosion breccias, which may include the one large grain described by the authors, are dominated by amorphous carbon, as identified by petrographic, EBSD and TEM studies. We have noted micro-inclusions of corundum, but not bauxite, in the matrix of the explosion breccias. The authors provide no evidence that bauxite, or any other Al-hydroxide, is present in the matrices. The EDS spectra given for “böhmite-like material” are equally consistent with corundum. We suggest that the report of böhmite /bauxite represents analyses of spots consisting of amorphous carbon plus micro-inclusions of corundum; this would be consistent with the apparently higher C content of the EDS spectrum. If böhmite actually is present, it may equally reflect partial alteration of corundum fragments. Whether this fine-grained material could be considered as an ash component is a semantic question.
6. “Osbornite from Carmel sapphire grains…as a rule are confined to glass inclusions”. This is not correct. The osbornite ‘skeletal crystals’ typically extend well outside the melt pockets, and occur as isolated inclusions in the corundum (as seen in 3D-μCT as well as 2-D slices; Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022). See also point (3) above.
7. “Griffin et al. (Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b) underline the non-stable composition of osbornite”. This is simply not correct; we have done no such thing. The cited reference presents analytical data and shows that they are consistent with experimental studies on the Ti–O–N system. There is nothing “non-stable” there.
8. “Bigger corundum aggregates contain silica glass”; “usually squeezed out between corundum grains”. This is misleading; the glasses in the melt pockets are calcium–aluminium–silicate (CAS) glasses (typically with 40–45 wt.% SiO2) with moderate levels of Mg, Ti and S, as well as high levels of REE and Zr. In some parageneses with cumulate-like structures the melts are indeed concentrated along grain boundaries, whereas in the more skeletal corundum aggregates the melts occur both along grain boundaries and as isolated inclusions. 3μCT images (Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a) illustrate this very clearly.
9. “.. ‘white breccia’ … as a source of Carmel sapphire”. This is misleading; nearly all investigated grains of Carmel sapphire have no traces of the white coating, and there is no evidence that they have been brecciated.
10. Zr-bearing glass as debris from “special laboratory glassware”. As noted above, the internal melts in the Carmel sapphire are high in Zr and the Carmel sapphire contains many Zr-bearing phases. Some corundum-aggregate xenoliths have thick rims of a Ba-rich or K-rich glass full of tiny crystals of zircon or baddeleyite. The high Zr contents reflect the fact that most of the melts left in the corundum aggregates by the time of eruption were highly differentiated. The spectrum in fig. 7d in the author's manuscript is very similar to those of the glasses in melt pockets in one type of Carmel Sapphire. It is unfortunate that these spectra were not accompanied by quantitative chemical analyses, which would have aided comparison with published data.
11. “.. terrestrial examples [of krotite and grossite] are questionable and require verification.” The authors might read our papers (Griffin et al. Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2020a) on the coarse-grained xenoliths of hibonite+ grossite+V0, which require a very H2-rich atmosphere and show the crystallisation sequence:
corundum + Liq → (low-REE) hibonite → grossite + spinel ± krotite → Ca4Al6F2O12 + fluorite. The occurrence of grossite in meteorites cannot invalidate the existence of these samples; their disaggregation during eruption is responsible for inclusions of krotite, hibonite and grossite in the breccias. These clearly did not result from the hydration of hypothetical precursor minerals.
12. “Glass fragments with inclusions of mullite”. Mullite does appear in some melt inclusions in the corundum aggregates. This is expected because the compositions of some residual melts fall into the liquidus field of mullite in CAS space at low P, and the crystallisation of mullite might be expected to occur during eruption.
13. “There is no mechanism for the simultaneous removal of Fe and Si from natural melt systems”. This is not correct. Reduction of an FeO-bearing silicate melt to f O2 = IW-5 (as documented by the abundance of Ti3+-bearing minerals in Carmel sapphire) will lead to immiscibility of Fe–Ti–Si silicides and supersaturation in Al2O3, leading to corundum crystallisation. Comparison of the analysed silicide-melt compositions with experimental data shows that this occurred at normal magmatic temperatures (1200–1400°C; Weitzer et al., Reference Weitzer, Schuster, Naka, Stein and Palm2008; Griffin et al., Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022). It is also worth noting that the presence of hydrogen can lower the melting points of metals and alloys by several hundred degrees (Fukai, Reference Fukai2005). Thus the 2000°C temperature of the industrial process is irrelevant to the natural case, although the observation that the silicide melts do sink to the bottom is a nice illustration of the process that we have proposed. We would appreciate a reference to the industrial process, which was not provided in the paper.
14. “formation from the melt requires extremely rapid quenching”. We agree; this is consistent with the very rapid eruption rate of the host basalts (pyroclastics) of the Carmel sapphire, which carry mantle-derived peridotite xenoliths (O'Reilly et al., Reference O'Reilly, Griffin, Dosetto, Turner and Van Orman2010)
15. “Therefore all new minerals [from] Carmel sapphire should be discredited by the IMA–CNMNC”. This is surely misguided (see below) and in any case is not appropriate to a journal article.
Finally, we must point out that a similarity between synthetic products (laboratory or industrial) and natural materials provides keys to the interpretation of natural processes (indeed, it is the raison d'etre of experimental petrology). However, such similarity cannot be taken as evidence for anthropogenic origin of the natural material. Duplication of mantle minerals in experiments does not imply that the minerals cannot exist in nature. This is the fundamental fallacy of the persistent criticisms by Litasov et al. (Reference Litasov, Kagi and Bekker2019a,Reference Litasov, Bekker and Kagib,Reference Litasov, Kagi, Voropaev, Hirata, Ohfuji, Ishibashi, Makino, Bekker, Sevastyanov, Afanasiev and Pokhilenkoc) and Ballhaus et al. (Reference Ballhaus, Wirth, Fonseca, Blanchard, Pröll, Bragagni, Nagel, Schreiber, Dittrich and Thome2017, Reference Ballhaus, Helmy, Fonseca, Wirth, Schreiber and Jöns2021) of the Mt Carmel work and many studies of similar mineral associations from numerous ophiolites by many researchers worldwide. Unfortunately, the experimental petrologists involved in this criticism have not followed scientific method; that would have included taking into account the overwhelming geological and geochemical evidence that makes an anthropogenic origin impossible, at least in the case of Mt Carmel (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
This evidence includes:
• The alluvial deposits in question are undisturbed Plio–Pleistocene palaeo-placers developed at the base of terraces that lie ca 10 metres above the current drainage of the Kishon River and are overlain by 4–10 metres of undisturbed finer-grained sediments (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
• Extensive sampling of known Cretaceous volcanic centres, which lie 50–350 metres above the Kishon River, shows that the vents collectively contain all of the minerals found in the alluvial deposits (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b), including moissanite and “Carmel Sapphire”.
• Miocene sedimentary deposits in the Yizre'l valley include carbonate-cemented, quartz-rich beach placers containing corundum, coarse-grained moissanite and peridotite-derived garnet+cpx+spinel (Roup et al., Reference Roup, Kamanovitch, Baykov and Toledo2009). Anthropogenic industrial processes were not available at this time.
• The occurrence of the corundum-moissanite suite in the primary volcanic centres, in deeply buried Plio–Pleistocene placers, and in lithified Miocene beach placers demonstrates that the sample material was formed and deposited long before human occupation and makes arguments for industrial contamination unsustainable (Griffin et al., Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reilly2021b).
• The detailed descriptions of parageneses (Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b, Reference Griffin, Toledo and O'Reilly2019b, Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a,Reference Griffin, Gain, Saunders, Cámara, Bindi, Spartà, Toledo and O'Reillyb, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022 and references therein) should lay to rest claims that such material was produced by lightning strikes (Ballhaus et al., Reference Ballhaus, Wirth, Fonseca, Blanchard, Pröll, Bragagni, Nagel, Schreiber, Dittrich and Thome2017).
• Cathodoluminescence imagery and Ti-zoning of corundum crystals (“Carmel Sapphire”) record growth from a silicate melt in an evolving open system at very low oxygen fugacity (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). This is not consistent with any industrial process known to us.
• In the Mt Carmel material, moissanite occurs in grains much larger than any common industrial abrasive (euhedral crystals up to > 4mm; Huang et al., Reference Huang, Xiong, Gain, Griffin, Murphy, Siryaev, Li, Toledo, Tomshin and O'Reilly2020). Aggregates of “Carmel Sapphire” and associated hibonite-grossite-vanadium assemblages are up to 2.5 cm across (Griffin et al., Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a, Reference Griffin, Gain, Camara, Bindi, Shaw, Alard, Saunders, Huang, Toledo and O'Reilly2020a).
The Mt Carmel mineral assemblages show significant differences to the synthetic materials produced and/or described by those arguing for a contamination origin (Litasov et al., Reference Litasov, Kagi and Bekker2019a,Reference Litasov, Bekker and Kagib,Reference Litasov, Kagi, Voropaev, Hirata, Ohfuji, Ishibashi, Makino, Bekker, Sevastyanov, Afanasiev and Pokhilenkoc).
• Tistarite, Ti2O3, in Carmel Sapphire does not have the high Al and Mg contents usually present in the synthetic material (Griffin et al., Reference Griffin, Gain, Adams, Huang, Saunders, Toledo, Pearson and O'Reilly2016).
• Synthetic ‘carmeltazite’ (Litasov et al., Reference Litasov, Kagi and Bekker2019a) shows a wide range in Zr content, requiring a high proportion of Ti4+ for charge balance. EELS analysis and single-crystal XRD analysis of the type carmeltazite from Mt Carmel show that all Ti is present as Ti3+ (Griffin et al., Reference Griffin, Gain, Bindi, Toledo, Camara, Saunders and O'Reilly2018a).
• Synthetic TiB2 in fused abrasives (Litasov et al., Reference Litasov, Kagi and Bekker2019a) contains significant levels of Al and F, which are absent in the TiB2 from Mt Carmel (Griffin et al., Reference Griffin, Gain, Saunders, Bindi, Alard, Toledo and O'Reilly2020b).
• The only EMPA-detectable minor elements in Carmel Sapphire are Ti, Si and Mg, however it contains trace elements (Sc, V, Cr and Zr), most of which are correlated with Ti contents ranging up to 2.5 wt.%. (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). In contrast, synthetic corundum produced from fused abrasives (Litasov et al., Reference Litasov, Kagi and Bekker2019a) contains much less Ti and no Mg, however it has significant contents of Fe, Ni, Mn, Ca, K, Th and P, which are absent or present only at low trace-element levels in ‘Carmel Sapphire’.
• TiC in the fused abrasives analysed by Litasov et al. (Reference Litasov, Kagi and Bekker2019a) contains no Fe, whereas that in ‘Carmel sapphire’ typically contains Fe at ca 1 wt.% levels, reflecting equilibrium partitioning with coexisting Fe–Si melts (Griffin et al., Reference Griffin, Gain, Saunders, Alard, Shaw, Toledo and O'Reilly2021a, Reference Griffin, Gain, Saunders, Huang, Alard, Toledo and O'Reilly2022).
• The most reduced of the Mt Carmel parageneses (corundum + hibonite + grossite + native vanadium) requires an atmosphere dominated by H2. To our knowledge, this does not occur in the production of alumina or Al2O3-based abrasives, because Al2O3 would be reduced to the metal, which is undesirable. However, it is a predictable result of the ascent and partial oxidation of COH fluids from a metal-saturated mantle (Griffin et al., Reference Griffin, Huang, Thomassot, Gain, Toledo and O'Reilly2018b, Reference Griffin, Gain, Huang, Saunders, Shaw, Toledo and O'Reilly2019a).
• The oxygen-isotope composition of Carmel Sapphire ranges from δ18O = +7 to +13‰, with a mean around +10 (Oliveira et al., Reference Oliveira, Griffin, Gain, Saunders, Shaw, Toledo, Afonso and O'Reilly2021). Natural corundum shows a range of δ18O from –27 to +23‰. In contrast, commercial synthetic-alumina products produced by a range of processes have δ18O = +23 to +32‰, reflecting the interaction of the material with air (δ18O = 23.9±0.3 ‰) during processing (Pack, Reference Pack, Bindeman and Pack2021). This difference is critical in distinguishing natural corundum derived from mantle/magma-related processes from a synthetic product.
• U-rich thorianite [(Th,U)O2] occurs with ziroite (ZrO2), mizraite-(Ce) [Ce(Al11Mg)O19] and Ti-sulfide in one melt inclusion in a Carmel Sapphire. EMPA Th–U–Pb age determination of the largest 1-μm thorianite crystal using a JEOL JXA-iHP200F Field Emission EPMA and the Probe for EPMA program yields an age of 13±2 Ma (Ma et al., pers com.), making it unlikely that the corundum host was produced by humans.
This evidence indicates that there is no scientific basis for the authors’ conclusion that “the contamination of geological samples with anthropogenic material has led to popularisation of biased views”. The responsibility for “popularisation of biased views” must lie solely with those who refuse to consider straightforward geological and geochemical evidence for the natural origin of the Mt Carmel samples.
Acknowledgements
This is contribution 1762 from the ARC Centre of Excellence for Core to Crust Fluid Systems (http://www.ccfs.mq.edu.au) and 1531 in the GEMOC Key Centre (http://www.gemoc.mq.edu.au).
Competing interests declaration
None of the authors have competing interests regarding this publication, and all authors have contributed to the authorship.