PGE-Enrichment in Magnetite-Bearing Olivine Gabbro: New Observations from the Midcontinent Rift-Related Echo Lake Intrusion in Northern Michigan, USA
Wolfgang Maier
Round 1
Reviewer 1 Report
Reviewer’s comments on “PGE-enrichment in magnetite olivine gabbro: New observations from the Midcontinent Rift-related Echo Lake intrusion in northern Michigan, USA” by Koerber and Thakurta
This is an interesting manuscript describing a new occurrence of PGE rich magnetite gabbro within a layered intrusion. The petrogenetic model is conventional but likely accurate, in my opinion. The topic of the paper is not new, but it is of considerable interest for the exploration and mining industry. I thus can recommend publication, but only after some relatively moderate revision, namely involving modifying some of the figures and improving the style of some sections in the discussion. I have annotated the pdf, but below are some key points:
L 51: give resource figures for Eagle and Tamarack.
L 59: whats the (potential?) thickness of the intrusion?
Fig. 1: show Sonju Lake.
L 101: more detail on the sufide mineralogy would be desirable.
Fig. 5: This is very confusing, please show column with continous meterage. You can then also add another figure where you highlight key mineralised zones.
Fig. 6: increase font of numbers, and add a column of PGE to see how these correlate with other elements.
L 166: this section could be a bit clearer: oxidation of magma causes Fe2O3 which, if it crystallises, should result in lowering of fO2 and thus sulfide liquid segregation?
L 224: check numbering of refs, there seem to be some errors.
L 229: surely no chromite?
From 233: style of writing is poorer from here on than in the preceding sections. I highlighted several sections in the annotated pdf that need rephrasing and clarification.
L 260: sulphide melt saturation (not sulphide saturation).
Comments for author File: 
Comments.pdf
Author Response
Reviewer 1: comments on “PGE-enrichment in magnetite olivine gabbro: New observations from the Midcontinent Rift-related Echo Lake intrusion in northern Michigan, USA” by Koerber and Thakurta
My responses are in blue.
This is an interesting manuscript describing a new occurrence of PGE rich magnetite gabbro within a layered intrusion. The petrogenetic model is conventional but likely accurate, in my opinion. The topic of the paper is not new, but it is of considerable interest for the exploration and mining industry. I thus can recommend publication, but only after some relatively moderate revision, namely involving modifying some of the figures and improving the style of some sections in the discussion. I have annotated the pdf, but below are some key points:
L 51: give resource figures for Eagle and Tamarack.
I have clarified the resource figures for Eagle and Tamarack in lines 45 and 52.
L 59: whats the (potential?) thickness of the intrusion?
The thickness of Echo Lake intrusion is not known. It is probably around 2 km. But it could be several kilometers thick like intrusions in the Duluth Complex. A continuous 1 km thickness has been confirmed from the exiting drilling data.
Fig. 1: show Sonju Lake.
I have shown Sonju Lake in Figure 1.
L 101: more detail on the sufide mineralogy would be desirable.
I have mentioned pyrrhotite, chalcopyrite and pentlandite as the most common minerals. There are minor occurrences of digenite and bornite. The overall textural nature of the grains has been mentioned.
Fig. 5: This is very confusing, please show column with continous meterage. You can then also add another figure where you highlight key mineralised zones.
I have removed the old Figure 5, and replaced it with Figures 5a, 5b and 5c. All these are drawn with continuous meterage. I have expanded the mineralized zone in Figure 5b.
Fig. 6: increase font of numbers, and add a column of PGE to see how these correlate with other elements.
I have increased the font size in Figure 6. The captions are now clear.
L 166: this section could be a bit clearer: oxidation of magma causes Fe2O3 which, if it crystallises, should result in lowering of fO2 and thus sulfide liquid segregation?
I have clarified this section and corrected some errors. When Fe2O3 forms in a magma it indicates a high ambient fO2 of the magma. However, when the precipitated Fe2O3 is removed from the system, the relative abundance of FeO increases again. This causes a lowering of fO2 and this might lead to the separation of a sulfide liquid.
L 224: check numbering of refs, there seem to be some errors.
I have corrected the error. I have changed the reference number from 54 to 53.
L 229: surely no chromite?
I have now clarified this issue. The Cr-enrichment is caused by high Cr-content seen in magnetite. There is no visible chromite in the assemblage. I have added a new table for magnetite composition, which shows the Cr2O3 content.
From 233: style of writing is poorer from here on than in the preceding sections. I highlighted several sections in the annotated pdf that need rephrasing and clarification
I have changed the paragraph substantially to make it clearer.
L 260: sulphide melt saturation (not sulphide saturation)
I have changed it to sulfide melt saturation.
Thank you very much. I appreciate all these comments. This has helped me to improve the manuscript substantially.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Dear the authors,
Please see attached file.
I hope these comments would improve the revised version.
best regards,
Reviewer
Comments for author File: Comments.pdf
Author Response
Response to Reviewer 2.
My responses are in blue.
This manuscript reported discovery of a new PGE-Cu enriched Echo Lake intrusion and preliminary petrology, mineralogy and whole rock data. The authors showed that the newly discovered intrusion has many similarities to other intrusions. However, data sets have not been provided yet (no whole rock data table), and petrogenesis is not well discussed based on their data set. The manuscript seems to me be a review of Cu-PGE rich intrusions.
I have added a new table: Table 1 with the whole rock chemistry of rock types.
The contents have significant interests for many aspects, but, as the authors have admitted by themselves (303-309 lines), the present manuscript is not many new inputs for the origin of the PGE-Cu concentrations and its-bearing tectonic settings. My decision is therefore acceptable with major revision. I left several comments below. I hope these ,comment would improve the revised manuscript.
The Echo Lake intrusion is a new discovery and very little is known about the entire extent and petrological character of this intrusion. Since the intrusion can be reached at a depth of 200 m from the surface, there is no surface outcrop. So, the work reported in this study is from drill core samples and preliminary information available from the exploration company.
We have prepared this short manuscript with a focus on PGE-enrichment in a specific unit. The purpose is to inform the exploration and mining geologists about a new PGE prospect and to highlight the occurrence of a new layered intrusive body in the Midcontinent Rift region, which is comparable with layered intrusions such as the Sonju Lake.
We believe, that the PGE-concentration in an oxide-rich layer which was crystallized from a differentiated magma is a very important observation. In a micro-scale, the location of fine sulfide mineral grains in close association with chromium-rich magnetite is a meaningful characteristic. As we have discussed in the article sulfide saturation was caused by rapid precipitation of oxide minerals from a differentiating magma.
However, we are still gathering new information about the intrusion and in the future we will publish new articles about the petrogenesis of this large layered intrusion. We shall keep in mind all the questions such as P-T-fO2 conditions of fractional crystallization, crystallization modeling from possible parental magmas, as pointed out in this review.
Major comments:
Analytical technique for whole rock chemistry.
Although the authors used a commercial lab for whole rock analyses, they should show more details for analytical details and also analytical accuracy and uncertainty, for instance, comparisons with standard samples. Otherwise, I cannot evaluate their data sets.
I have attached quality control certificates for our data from the commercial lab. These include certificates for all the methods of bulk data collection.
Figure 5: Depth scale is not proportional (around 1000-1100 m are detailed) in Fig. 5. I recommend the authors to prepare diagram showing element concentrations vs depth scale, then prepare a new diagram focusing on only few sequences including the PG Erich magnetite layer.
I have removed the old Figure 5, and replaced it with Figures 5a, 5b and 5c. All these are drawn with continuous meterage. I have expanded the PGE-rich in Figure 5b.
Discussion, 4.1. Sulfide-saturation in mafic-ultramafic magmas This subchapter is not discussion but review from other studies.
4.2. Layered and conduit-type intrusions associated with the MCR
This subchapter is also review rather than discussions……
I started the discussion chapter with two subchapters: 4.1: Sulfide-saturation in mafic-ultramafic magmas and 4.2: Layered and conduit-type intrusions associated with the MCR. The intent was to first discuss the existing knowledge about layered intrusions in the Midcontinent Rift region and to put the Echo Lake intrusion in the proper geochemical and regional context.
Then in the subsection 4.3: Inferences form Echo Lake intrusion, I have used all this information plus data from this study to create a model of origin of PGE mineralization. So, the first two subsections are reviews, based on which the final interpretation is made.
4.3 Inferences form Echo Lake intrusion
The authors suggested the correlations between major & minor elements and PGEs and Cu only in Fig. 6, but these correlations are not clear for me only from Fig. 6. Please prepare several key variation diagrams for discussions.
We showed the variation diagram for Cu, Ni and PGE in Figure 5 (a, b, c). In Figure 6, the variation diagrams with respect to MgO, FeOT, Ti and Cr are shown. This choice is based on the proxies for degree of differentiation of the magma and the key elements which show direct relationship with PGE and Cu.
I can show other variation diagrams of the layered units but did not include those in this article to keep the article short. We shall include more variation diagrams in an article on the detailed petrogenesis of this intrusion.
Please explain why both chromite and ilmenite are locally concentrated in a specific horizon in the intrusion, because Cr is compatible elements suggesting that chromite is generally thought to be early crystallization phase than ilmenite.
I have corrected this in the revised manuscript. As pointed out by reviewer, chromite and ilmenite cannot form together in the same horizon. The high Cr2O3 content in certain oxide-rich intervals is not caused by chromite but by high Cr-contents in magnetite. Magnetite co-exists with ilmenite. For reference, I have added the composition of magnetite in Table 4. Free chromite is not seen in the mineralized intervals.
23-236 lines: The lithological succession of the Echo Lake intrusion is similar to tholeiitic
layered intrusions as above and like all Skaergaard-type deposits, the origin of the PGEenrichment in Echo Lake intrusion is consistent with the attainment of sulfide saturation in a high-aluminum olivine tholeiite (HAOT) magma by fractional crystallization in a closed system [8, 5].
The lithology and relationships between the studied intrusion and other intrusions are similar. This is OK. But the similarly does not directly mean the origin is the same. Please discuss more clearly what is lines of evidence that the studied intrusion was derived from a high-aluminum olivine tholeiite magma by fractional crystallization?? Please discuss more carefully P-T-fO2 conditions on fractional crystallization and its derived chemical change in fractionated melt, for examples. Otherwise, this paper is not scientific results.
I have included more information about the high-aluminum olivine tholeiite magma and its relevance to the Echo Lake intrusion. I agree that petrological similarity does not necessarily mean similar origin. So, I have mentioned this as a possibility that is consistent with the observed petrogenetic and geochemical characteristics. I have explained this in more detail based on this review.
As mentioned above, I shall give detailed descriptions of the petrogenetic model based on P-T-fO2 conditions and fractional crystallization of magma, in a future article. This article will explain the origin of the intrusion in broader detail.
251-253 lines: Owing to the highly incompatible nature of PGE in silicate minerals, particularly Pt and Pd, the concentrations of these metals in the magma increases by progressive fractional crystallization in a closed system. Pt and Pd might be incompatible elements but other PGEs such as Ir, Ru, Rh (and Os) are compatible.
PGE are incompatible in silicate minerals in general, but this varies slightly between PPGE and IPGE. IPGE are slightly more compatible in silicate magma, but the degree of compatibility is so small that, it does not matter much. So, it is true that the concentration of PGE increases by fractional crystallization of silicate minerals. This was pointed out to me by the editor also.
268-269 lines: An HAOT-type magma with ~13 wt.% FeO [56] began crystallizing olivine, plagioclase and smaller quantities of pyroxene. Why ~13 % FeO? How about other elements? That melts can crystallize olivine to form troctolite? Why chromite crystalize such an evolved melt?
I have mentioned more details about the HAOT magma in lines 285-294. I have added more compositional data and the calculated abundances of olivine, plagioclase and clinopyroxene by the crystallization of HAOT in a closed system. I have deleted “~13 wt% FeO” in line 268.
The Cr2O3 enrichment was caused by Cr-rich magnetite, not chromite. I have corrected this in the revised manuscript.
288-291 lines: This caused localized zones of sulfide saturation and the formation of tiny droplets of immiscible sulfide liquid. These immiscible droplets formed at isolated pockets, sequestered the chalcophile elements such as Pd, Pt and Cu from the surrounding melt mineralized horizon.
Liquid immiscibility of sulfide-rich melt is Ok, but any petrological evidence?? Is Fig. 4 only potential evidence for liquid immiscibility? Can this occurrence (Fig. 4) be explained oxide-sulfide crystallization from fractionated melt rather than from sulfide-rich liquid?
Yes, as in Figure 4, the small sulfide grains surrounded by, or included within silicate minerals indicate immiscibility of sulfide liquid. We have many other photomicrographs of such textures, but I chose only a few for this article.
For the rim like chalcopyrite grains around magnetite, the sulfide minerals were formed by the crystallization of small droplets of sulfide-liquid. The sulfide liquid was locally produced by the crystallization of magnetite from a fractionated melt.
Table for whole rock chemistry??
I have added a table (Table 1) for whole rock chemistry for the major rock types.
Minor comments:
Significant number for tables.
For example, Table 1. Ni ppm in olivine: Ni ppm in some olivine grains is shown up to 2 decimal places. It might be better to consider significant number for analytical data.
Other example, Fo of olivine is also shown up to 2 decimal places… but it might not be required. Fo 61.61 is different from 61.6??
I have changed the data tables for Fo, Ni in olivine and Ab, An, Or for plagioclase to one place of decimal.
438 line: Publication age for Carroll & Rutherford
I have added the year, 1988 for Carroll and Rutherford.
I hope these comments would improve the revised version of the manuscript.
Thank you very much. I appreciate all these comments. This has helped me to improve the manuscript substantially.
Best regards,
Author Response File: Author Response.pdf
