Log InRegister
Home PageAbout MindatThe Mindat ManualHistory of MindatCopyright StatusWho We AreContact UsAdvertise on Mindat
Donate to MindatCorporate SponsorshipSponsor a PageSponsored PagesMindat AdvertisersAdvertise on Mindat
Learning CenterWhat is a mineral?The most common minerals on earthInformation for EducatorsMindat ArticlesThe ElementsBooks & Magazines
Minerals by PropertiesMinerals by ChemistryAdvanced Locality SearchRandom MineralRandom LocalitySearch by minIDLocalities Near MeSearch ArticlesSearch GlossaryMore Search Options
Search For:
Mineral Name:
Locality Name:
The Mindat ManualAdd a New PhotoRate PhotosLocality Edit ReportCoordinate Completion ReportAdd Glossary Item
Mining CompaniesStatisticsUsersMineral MuseumsMineral Shows & EventsThe Mindat DirectoryDevice Settings
Photo SearchPhoto GalleriesNew Photos TodayNew Photos YesterdayMembers' Photo GalleriesPast Photo of the Day GalleryMineral Photography

Yukon Phosphate Update 2020

Last Updated: 7th Mar 2020

By Richard Gunter


Mr. Richard Gunter M.Sc., P. Geo (ret.)
5493 Cedarcreek Drive
Chilliwack, British Columbia
V2R 5K4

Table of Contents


The phosphate minerals from the Rapid Creek Formation, Yukon Territory, Canada are some of the most distinctive mineral specimens on the market. Well crystallized examples of Lazulite, Whiteite, Augelite, Kulanite, Gormanite and other rare species grace mineral collections world-wide. They all come from small veins in the phosphorous-rich iron formation in the northeast corner of the Yukon Territory.

Robinson et al. (1992) in their Mineralogical Record Special Issue on the Yukon Phosphates compiled a complete and detailed account of the phosphate occurrences with the chemical analyses and mineral characterization completed by 1992. Since that date there has been considerable mineral analyses and the characterization of several new mineral species. This update is designed to bring the database for the Yukon phosphate sequence up to date.


Since the publication of the “Yukon Phosphates Special Issue (Mineralogical Record v. 23 no. 4) by Robinson et al. (1992) there has been considerable examination of various Yukon phosphate species. The Arrojadite Group that was outlined, but not finalized, in the Special Issue has been examined and “Bobdownsite” has been approved and subsequently discredited as a separate mineral phase. Several associations have been investigated in greater detail such as the occurrences of the phosphate nodules and their distinctive mineralogy.

The regional geology for the Rapid Creek and Blow River area has been outlined in the Special Issue and has not changed. A detailed look at the local geology (Yeo 1992) was published contemporaneous with Robinson et al. (1992) (the Special Issue). Yeo (1992) does not reference Robinson et al. (1992) and Robinson et al. (1992) does not reference Yeo (1992).

Individual localities (Cross Cut Creek Area A; locality 1 for example) have been superseded by more exploration of the surrounding river banks and mountains by mineral collecting parties and the finding of different parageneses outside the outlined localities, such as the Whitlockite (“Bobdownsite”) locality. An example of the locality change is on Cross Cut Creek on which Robinson et al. (1992; Figure 4) has three noted localities. Detailed mapping of Cross Cut Creek by Yeo (1992; Figure 1) has seven noted occurrences of “epigenetic phosphate minerals” over a similar length of creek. One of the Yeo (1992) occurrences has been detailed mapped as a 92 meter thick stratigraphic section of the Rapid Creek Formation on the western end of Cross Cut Creek where he has located six sub-occurrences of “epigenetic phosphate minerals” in a variety of sedimentary host rocks. Yeo (1992) did not identify the mineralogy of the “epigenetic phosphate minerals” or note any differences in mineralogy between his occurrences, as Robinson et al. (1992) did.

The thesis by Robertson (1980) is the first major analysis of the locality and has much compositional data that was used and supplemented by Robinson et al. (1992). Robertson’s localities are also fewer than in Robinson et al. (1992) but Robertson provides maps of mineral abundances that are invaluable. An update to some of the Yukon phosphate geology and several of the individual phosphate phases is Tomes et al. (2013). She did not provide maps for the phosphate localities but provided photographs of the phosphate-iron formation outcrops missing in the older literature.

Robertson assumed that all of the phosphate samples were H2O and OH dominant and did not analyze for F. Other analyses of the Yukon phosphate samples, especially the detailed analytical work that has been undertaken on the Arrojadite Group (Camara et al. (2006), Chopin et al. (2006) and Tomes et al. (2018)) indicate that F is present in small quantities within the hydrous phosphates. The characterization and subsequent discreditation of “Bobdownsite” (Tait et al. (2011) and McCubbin et al. (2017) respectively) was based on the presence or absence of an F atom in the crystal structure of the Whitlockite phase found in the Yukon phosphate suite.

The combination of an unusual host rock in the Mesozoic Rapid Creek phosphatic iron formation, Tertiary brittle fracturing and the lack of significant post-Tertiary weathering has produced the veinlets of crystallized Mg-Mn-Ba phosphates that have been collected and distributed around the world.

Phosphate Iron Formation Geology

The Rapid Creek Formation is placed in the uppermost Lower Cretaceous (Western Canadian Sedimentary Atlas, 1984) and is thus related to the Early to Mid-Cretaceous Columbian Orogeny. The Rapid Creek Formation is an unusual phosphorite-carbonate iron formation sequence that appeared to form at very high latitude during the mid-Cretaceous (Yeo, 1992); a rarity according to Cook and McElhinny (1979). All of the Mid-Late Cretaceous sedimentary sequences in Alberta-British Columbia and the Yukon have been deposited at high paleo latitude. The southernmost group of coal-bearing beds in the Coalspur area of central Alberta was deposited at 60o N latitude. Further south in British Columbia the coal bearing horizons are slightly older than the coal-bearing strata in central and northern British Columbia with the Lower Jurassic to Lower Cretaceous Kootenay Group being the main coal-bearing unit and the overlying Lower Cretaceous Blairmore Group being largely conglomerate and sandstone; thus the Rapid Creek Formation was formed during subsidence.

The phosphate mineralogy of the Rapid Creek Formation is not (now?) dominated by minerals of the Apatite Group but by satterlyite and gormanite (Robertson (1980); Tomes, (2013)). Thus it has a different mineralogy than all of the other phosphorite occurrences. The iron phosphate mineralogy of the Rapid Creek Formation is not characteristic of cold water phosphorite deposition as glaciogenic phosphorite in the Permian Karroo Supergroup of South Africa is carbonate-fluorapatite (Buhmann et al. 1980). The unusual satterlyite and gormanite mineralogy is not localized as it appears in all of the Rapid Creek Formation outcrops in the Big Fish and Blow River watersheds and Boundary Creek (Figure 1).

Figure 1: Location Map for the Yukon Phosphate Mineralization from Robertson (1982)

Figure 2: Cross-cut Creek in 2012. (Tait pers comm; 2019)

Figure 3: Waterfall on Cross-Cut Creek in 2012; Mr. Bob Beckett is scale (Tait pers comm; 2019)

The unusual mineralogy of the Rapid Creek Formation is reflected in the noduals which are locally abundant in the lower part of the strata over a wide area. There are no macroscopic indications (bleached halo etc.) that the mineralogy of the noduals is not in equilibrium with the surrounding phosphatic shale. The noduals themselves, whether they are composed of satterlyite, maricite, wolfeite or other phases, do not have a reaction rim on the outside of the nodual and the prismatic crystals of the interior minerals grow to the edge of the nodual. David Joyce says: At certain places in the Yukon phosphate deposits, rare phosphate minerals occur in flattened nodules in the iron-phosphate rich shales. Some of these species are very rare. How can you tell the species apart? Wolfeite = cinnamon brown, lustrous; Satterlyite = resinous yellow; Maricite = milky pink/greyish dull lustre, often with metavivianite giving blue colour in between crystals of maricite. In section, maricite looks like dinosaur bone. Wicksite = black or very dark green. Non-nodular spays of satterlyite also grown in the phosphatic shale without reaction rims (Tomes (2013).

Some, but not all, of the noduals are pseudomorphed after fossils (Robertson, 1982). The majority of the interior minerals in the noduals are non-Ca bearing (Tomes, 2013) so the Ca from the fossil shells would not have a control on the mineral precipitation. Thus it appears that the mineralogy of the noduals is in equilibrium with the phosphate mineralogy of the surrounding phosphatic shale. Coleman and Robertson (1981) in the characterization of nahpoite says: A brief study of compositions of nodules observed in situ suggests that 39% are composed primarily of pyrite, 58% of wolfeite, 2% of satterlyite and l% of maricite.

Flattened nodules in phosphatic shale

Rapid Creek, Dawson mining district, Yukon, Canada
Image 1: Angular, flattened noduals in shale with no reaction rims and localized disturbance of the bedding within the shale. The nodual in the centre, above the notebook, has the characteristic shape and alteration of a satterlyite nodual, some of which contain pyrite.(Tait pers comm; 2019)

Image 2: Blue nodual of Baricite looking downward on the bedding plane of the shale. In this view, perpendicular to the above view, the shale is not deformed by the growth of the nodual. (Tait pers comm; 2019)

Image 3: Slightly rusty, sub-spherical, isolated nodual showing no reaction rim. The noduals can vary from loonie to palm size and are often isolated in the phosphatic shale. (Tait pers comm; 2019)

Image 4: Unweathered interior of a split Satterlyite nodual with radial Satterlyite on the right of the photograph and slightly rusty pyrite on the left side of the photograph. The outside of the nodual is the same colour and shape as the in-situ nodual in Figure 1.

The crystallization of the macroscopic crystals of Satterlyite must have postdated the consolidation of the shales, given the disruption of the shale layers surrounding the Satterlyite aggregates (Image 25) and the noduals (Image 1). There is no evidence of the growth of the phosphate minerals on the outside of any of the noduals (Images 2 and 3) so the growth of the Satterlyite, Wolfeite etc. in the interior of the noduals must have occurred after the noduals formed. It is possible that the growth of the nodular and non-nodular Satterlyite was simultaneous and the mineralogy of the noduals responded to highly localized pH/Eh conditions.

The mid-Cretaceous Rapid Creek Formation sedimentary sequence was subjected to Tertiary (Eocene) folding and thrust faulting, as are many Cretaceous sedimentary sequences along the length of the eastern Canadian Cordillera. The folding and faulting resulted in the creation of extension veins, similar to many others in the eastern Cordillera of British Columbia and the Yukon. The Rapid Creek Formation sediments are relatively calcium-poor and silica poor so the veins are not flooded by Calcite or Quartz.

Rod Tyson, who has visited the area several times, said (pers comm., August 8, 2019): Hi Richard, Quartz is common – not sure if it is ubiquitous, never paid that kind of attention. There is usually a first gen quartz that is under a lot of the phosphates or co-eval. There is also a second gen quartz that makes Herkimer like pieces – these are much more rare.

Figure 4 “S” shape veinlet in shale illustrating the typical vuggy open-space filling nature of the veinlets.

Within the sedimentary sequence there are repeated layering of contrasting grain size, from very fine-grained siltstone to conglomeratic deposits. Yeo (1992) considers this to be deep water, turbidite based, sedimentary sequence with the silt and sand sized units corresponding to the classic turbidite layering and the conglomerates representing slumped deposits at the active base of the turbidite layers. This type of homogenous turbidite layering would tend to mix and average any chemical variations in the pre-turbidite layers. Thus the division by Robertson into four phosphate units, based on local chemical variability within the sediments, would be unlikely.

Cross-section of Rapid Creek Formation along Crosscut Creek

Crosscut Creek, Kulan Camp, Rapid Creek, Dawson mining district, Yukon, Canada
Figure 5: Outcrop on Cross Cut Creek illustrating the interbeds of sideritic ironstone (thick beds) and phosphatic shale (thin beds).

Note the repetition of virtually identical phosphatic shale and carbonate iron formation with little in the way of sedimentary structure. This is consistent with the deposition as deep-water mixed turbidite sediment which will tend to result in large-scale homogeneity in the chemical composition of the sediments.

Figure 6: Localized conglomeratic slump deposits (orange coloured) from the base of the tubidites Tait (2012; pers comm.)

Yeo (1992) indicates that the conglomeratic slump deposits are scoured into the sedimentary section only at the base and not the top; thus they are deep water turbidite rather than a shallow water sedimentary section. The scour can be seen as an unconformity with the lower shale layers the base of the orange layer.

The sediments as a whole contain sufficient iron and phosphorous that they were examined as a potentially economic resource by Welcome North Mines (Brock, 1975) who commented on the appearance of the phosphate minerals though that was not their focus. Their two samples gave a mineralogical analysis of:

Table 1: Mineralogical Content of Rapid Creek Shale Samples (Brock, 1975)
Sample 108-YA-1 108-YA-4
% apatite 05.94 10.75
% kaolin 08.14 06.88
% pyrite 00.06 02.12
% siderite 55.95 31.06
% quartz 11.28 06.46
% non-crystal 18.64 42.73

The high percentage of non-crystalline component combined with the low chemical content of SiO2 not accounted for by quartz and kaolin: -0.18% and-016% (within the margin of error but essentially 0%) respectively makes it very unlikely that the non-crystalline component represents clay minerals. There is a large difference in the measured siderite content in the samples: 55.95% (34.69% FeO) and 31.06% (19.26% FeO) for a negligible increase in the Total Iron content: 38.61% and 34.12% respectively. The high chemical P2O5 content of 7.85% and 20.00% for the two samples is far higher than can be accounted for by the mineralogically determined apatite percentages: representing 2.51% and 4.54% P2O5 respectively. (It is possible this is the mineral Santabarbaraite (characterized by Pratesi et al. 2000)). The CaO chemical content of 3.31% and 5.59% correlates to 5.96% and 10.06% apatite; approximating the determined apatite content of 5.94% and 10.75% respectively There is also a high L.O.I. for the samples 21.08% and 14.75% respectively that will influence the chemistry.

The plotting of the chemistry on the measured mineral content of the two shale samples indicates that a significant portion of the phosphorous in both samples and a significant portion of the Total Iron in sample 108-YA-4 report to the non-crystalline portion of the mineral analysis. The CaO and Si02 content of both samples is accounted for by measured percentages of apatite, quartz and kaolin. The amorphous iron phosphate is probably the precursor to the late-stage phosphate minerals.

One of the unexplained occurrences in the Rapid Creek Formation is the presence of sufficient mobile Fe, Mg, Ca and CO3 to form abundant co-existing siderite, lazulite and fluorapatite in the late-stage veins but not calcite or dolomite. Significant volumes of apatite and siderite occur in the Rapid Creek Formation shale mineralogy (Table 1) but no calcite or dolomite is present in the shale or in the noduals. Mobile silica and aluminum are abundant as well but there are no hydrous silicates formed either in the veins or the noduals; though significant kaolin occurs in the shale.

Metamorphic temperature from fluid inclusions is discussed in Robinson et al. (1992) and Yeo (1992). There appears to be no source of hydrothermal fluids in the Rapid Creek Formation that was not present in the regional mid to late Cretaceous sedimentary strata on the east slope of the Canadian Cordillera. Since there are significant bituminous and sub-bituminous coal measures in the strata south of the Rapid Creek Formation there are strong indications of a relatively low upper limit to the temperature of any regional fluid flow.

Similar veinlets to the phosphate-bearing ones at Rapid Creek occur in coal measures south of the Rapid Creek Formation. The 6 cm x 10 cm cavities have linings of platy Calcite-Pyrite (Wildhay, Alberta) and rhombohedral Calcite-Quartz (Babcock Mountain, British Columbia). In both cases the veinlets occur in response to fracturing of a sandy layer within a plastically deformed shale sequence. The elongation of the veinlets is at right angles to the bedding of the sandy layer. There is no deformation of the surrounding sediments immediately adjacent to the veinlets so they do not represent fault filling. The mineralized veinlets occur as part of a swarm of parallel veinlets.

The enclosing coal measures were typical sandstone-shale-coal sequences that have been folded and faulted. The faulting represents an over thrust sequence with the thrusting from a westerly direction. The over thrust faulting was not confined by overlying strata and the low-pressure zones are often open veins with crystal lined cavities. The low-pressure zones can be areas of ductility contrast as at Babcock Mountain or they can be mechanical bending of a more competent, carbonate cemented, sandstone layer as at the Wildhay Occurrence.

In both cases the Calcite was the first phase and micro-crystals of cubic Pyrite at Wildhay and 2 cm crystals of sceptered Quartz at Babcock were the last phase to form. There is less Pyrite present in the coal measures on Babcock Mountain than there is in the Wildhay area so the filling of the veinlets represents locally derived components. There are no reports of the mobilization of titanium minerals and the paleo placer deposit at the Burmis Occurrence does not contain evidence for titania mobility. Thus the presence of minor Anatase in the Rapid Creek phosphate paragenesis may indicate a slightly higher temperature deformation on the northern end of the Cordillera.

There is only one occurrence of volcanic rocks within the eastern Cordillera, represented by the Crowsnest Pass analcime-bearing extrusive volcanics. Thus there was no intrusive heat from volcanism in any of the deformed chemical sequences. The Crowsnest Pass volcanics are compositionally very distinct with abundant Ti-enriched Andradite and primary Analcime so the presence of these volcanics in the parental rocks of the sedimentary sequence would be noted by the unusual chemistry of the detrital minerals within the sandstones and shales.

Mineralogical Changes since 1992 Special Issue

The many collecting expeditions undertaken to the Yukon Phosphate locality since the Special Issue’s publication in 1992 have uncovered and sampled numerous additional phosphate-bearing veins. Only a few new mineral species have been identified; “Bobdownsite” and Arrojadite-(Ba,Na) (Tomes (2011);Tomes et al. (2018)) but much new chemical and structural data (RRUFF database and others) has been produced. Some phases new to Rapid Creek such as Xanthoxenite (Reiner Meilke), Matioliite (Alfredo Petrov) and Metavauxite (Pavel Kartashov) have been reported in Mindat.org with supporting data but with only vague locality data. The Alluaudite Group (Hatert; 2006) and the Alunite Group (Dzikowski et al.; 2006) have been reclassified in part using samples and data from the Yukon Phosphates.


The Whitlockite-“Bobdownsite”-Whitlockite controversy is the major change in the Yukon Phosphate paragenesis since 1992. Whitlockite was noted briefly in Robinson et al. (1992). “Bobdownsite” was characterized from the same Big Fish River locality by Tait et al. (2011). A good description of Whitlockite paragenesis is in Tomes (2013).

"Bobdownsite"-Whitlockite vein in situ. Dr. Ron Peterson for scale.

Rapid Creek, Dawson mining district, Yukon, Canada
Figure 7: “Bobdownsite”/Whitlockite vuggy vein with Quartz; cutting across the stratigraphic section; Dr. Ron Peterson is on the outcrop. (Tait pers. com.; 2019)

An article discrediting “Bobdownsite” as equivalent to Whitlockite is McCubbin et al. (2018). RRUFF still considers “Bobdownsite” a valid species with posted analyses that do not match the analyses from McCubbin. RRUFF “Bobdownsite”: 1 average sample analysis; n=25
(Ca8.76Na0.24)Σ=9(Mg0.72Fe3+0.13Al0.11Fe2+0.04)Σ=1(P1.00O4)6(P1.00O3F1.00); Fe2+ and Fe3+ by charge balance

The Whitlockite veins have a complex mineralogy that has not had all of the phases chemically analyzed. In addition to the colourless, platy Whitlockite there are elongate blades of Gormanite and thin, warped platy aggregates of Kulanite that grow at right angles to the Gormanite blades. The earliest phase in the paragenesis is a member of the Arrojadite group in a 1 mm thick layer comprised of an aggregate of tan coloured, altered crystals mixed with colourless quartz crystals on an unaltered appearing shale matrix. The latest phase in the paragenesis is micro-crystalline white to light tan coloured aggregates of Nahpoite that grows on most of the earlier phases.

An example of the paragenesis is:

Image 5: Whitlockite Paragenesis Big Fish River; sample 5 cm in length.

Image 6: Whitlockite/”Bobdownsite” and other phosphate phases at the site of the “Bobdownsite” vein. (Tait pers com; 2019)


A new member of the Arrojadite Group has been described from the Whitlockite paragenesis (Tomes (2011) and Tomes et al. (2018)). She says: At Big Fish River, it is associated with “bobdownsite”, whitlockite, gormanite, augelite, kulanite, collinsite, siderite and quartz. The crystals are tabular and colourless to slightly orange in colour. An illustration of the new mineral (Image 7) is associated with Augelite.

Image 7: Arrojadite-(Ba,Na) associated with white Augelite; photograph from Tomes (2011)

Alluaudite Group

The Alluaudite Group is confined to the noduals at Big Fish River. These noduals contain: Satterlyite, Wolfeite, Baricite, Maricite, Wicksite, Alluaudite, and Nahpoite. Data on the distribution of the noduals is in Tomes (2013) as there was not much paragenetic data in Robinson et al. (1992). The RRUFF database has one Arrojadite Group phase from the noduals:
RRUFF Ferroalluaudite: (nodual with maricite) 1 average sample analysis; n=8

Alunite Group (Gorceixite and Goyazite)

The Alunite Group has been reorganized by Bayliss et al. (2010) and the two members of the group that occur at Rapid Creek are Gorceixite and Goyazite. Both species have their best expression in the world at this locality.

There have been no more new finds of Gorceixite since 1992. All Gorceixite samples from the literature appear to be from the same find described by Robinson et al. (1992). Dzikowski et al. (2006) has an article with detailed analyses of the Gorceixite from Rapid Creek with mineral chemistry and structural analyses indicating this is an unusual member of the Alunite Group. The variability in chemistry noted in Robinson et al. (1992) was not noted in Dzikowski (2004) or Dzikowski et al. (2006). Dzikowski (2004) used a sample from Rapid Creek for her analysis of the crystal chemistry of Gorceixite. In her three spot analyses of the sample she had (in apfu): Ba 1.011; 1.008 and 0.949 and Sr 0.013; 0.014 and 0.012. The RRUFF database does not have analyses of Gorceixite from Rapid Creek.

A more common member of the Alunite Group at Rapid Creek is Goyazite. The RRUFF database has one Goyazite sample posted but has done no analyses. A fairly common paragenesis of Goyazite discovered since 1992 is with pseudocubic, slightly pink Goyazite, colourless, octahedral Wardite and short prismatic, colourless Fluorapatite. An illustration of that paragenesis is:

Image 8: Goyazite-Wardite-Fluorapatite Paragenesis; Rapid Creek. Sample is 4.5 cm long.

There have been other Goyazite finds without Wardite and Fluorapatite but this appears to be the most common paragenesis. The collector of this sample, Rod Tyson, indicated the locality was “not too far from Stoneman Camp”.

Arrojadite Group

The complexity of the Arrojadite Group, hinted at in Robinson et al. (1992), has been the subject of considerable study since then. Camara (2010) and Chopin (2010) in companion articles re-organize the Arrojadite Group and give analyses of the new members. The RRUFF database notes two different Arrojadite group minerals at Rapid Creek; one from the nodules with chemistry and one from the veins:

RRUFF Arrojadite: 2 samples; 1(crystal) has no analysis and 2 (from the nodual locality) has an average sample analysis; n=10
(Na1.80K0.20)Σ=2Fe2+1.00(Ca0.90Na2.10)Σ=3(Fe2+10.71Mg1.40Mn0.89)Σ=13(Al0.90Ti0.05Fe3+0.05)Σ=1(P1.00O4)11(P1.00O3OH)((OH)1.15F0.85)Σ=2; (OH) by stoichiometry and charge balance

An example of Arrojadite-(K,Fe) from the veins is:

Image 9: Arrojadite-(K,Fe) sample; Gianfranco Ciccolini (2019) photograph in Mindat.org.

Additional Finds and Data since 1992


Not much new since Robinson et al. (1992). E-minerals have it occurring with Brazilianite and Kulanite but that is unconfirmed. The RRUFF database has no Anatase from Rapid Creek


Augelite has been found in more localities than 1992; some cavernous. CMN staff brought back Augelite samples from collecting @ locality 5. Augelite almost always occurs alone in the vein cavities without many of the other Mg-Fe phases.
RRUFF Augelite: 2 average sample analyses posted; n=8.
1) Al2(P1.00O4)(OH0.98F0.02)3 2) Al2(P1.00O4)(OH)3

Many Augelite photos have been posted from Mindat.org and commercial mineral websites such as David Joyce. An unusual Augelite is a 5 cm x 3 cm sample from Rapid Creek where the 1 cm light green Augelite crystals are locally cavernous. This is a very unusual texture that does not show up well in photographs. I do not know how the texture was created but the more cavernous Augelite crystals have less brown inclusions within them. The intact Augelite crystals have brown inclusions that are not crystalline, but botryoidal, yet the cavernous portions of the Augelite crystals are always bounded by negative crystal faces.


The E-minerals website has brown Baryte crystals associated with Whiteite; Baryte from Rapid Creek is mentioned briefly in Robinson et al. (1992).


Brazilianite is briefly covered in Robinson et al. (1992). New data from Corbic et al. (2011) for Brazilianite morphology contradicts Robinson et al. (1992) on Brazilianite morphology. The RRUFF database has no samples for the Rapid Creek Brazilianite.


A brief overview of the Childrenite occurrence in Rapid Creek was in Robinson et al. (1992). More chemical data was provided by Pavel Kartashov who said in an e-mail to me on April 6, 2012: investigated by me Childrenite-Eosphorite crystals from Yukon were transparent, gemmy, tea-like colour without visible zonation. This is one of the less common of the Yukon phosphate species. According to Robinson et al. (1992) the Childrenite-Eosphorite samples are widespread but do not produce good examples. They do not describe the high iron Locality C samples which form a different Childrenite-Eosphorite suite than the Rapid Creek Mn-rich paragenesis.

More recently good Childrenite samples have come from the Big Fish River area. The bulk of the Big Fish River Childrenite find is later than 1992 so there is limited data on its paragenesis or chemistry. The associated minerals often do not help as they are widespread as well. All of the Childrenite photos in the literature look alike and are probably from this sub-locality, though detailed locality labels are rare.

There are no Vauxite group minerals in this paragenesis though Gormanite is quite common, the Big Fish River area being the best Gormanite locality. The Big Fish River paragenesis is missing many of the high Mg or Mn phases so it may have been significantly more iron-rich, or had more iron available, than the Rapid Creek phosphates. In many ways this paragenesis is transitional to the phosphate nodules. The RRUFF database has no Childrenite samples from Rapid Creek.


Collinsite at Rapid Creek is essentially a one-sub-locality mineral that has been described in detail in Robinson et al. (1992). New chemical data from the RRUFF database is:
Collinsite: 1 average sample with light and dark phases on the grey scale of the backscatter image; n=10
Dark phase is: Ca2.00(Mg0.63Fe2+0.34Mn0.03)Σ=1(P1.00O4)2·2H2O ; The light phase is: Ca2.00(Mg0.58Fe2+0.38Mn0.04)Σ=1(P1.00O4)2·2H2O

Image 10: Backscatter image Collinsite, Big Fish River; RRUFF (2019) database photograph.

The slight increase in Fe+2 in the lighter phase seem to be crystallographically controlled as the light phase occurs as lenticular inclusions in the darker phase (Image 10). The euhedral Collinsite crystals from Rapid Creek are lense-like in outline and often have complex, non-planar, faces (Image 11). The lense-like compositional zones seen in Image 10 appear as multiple, intergrown, lath-like sub-crystals in each of the bladed Collinsite crystals noted in Image 11.

The Collinsite paragenesis is unusually restricted with only a few other minor phases in the Collinsite veins. Quartz and Siderite, almost universal in the other phosphate veins, do not occur here. The Collinsite crystals in these samples grow perpendicular to the unaltered Rapid Creek Formation matrix and often span the entire vein.

Image 11: Euhedral Collinsite on matrix; Big Fish River


The Fluorapatite descriptions in Robinson et al. (1992) have been supplemented by more data from Tomes (2013). Fluorapatite from Rapid Creek has not been analyzed in the RRUFF database. Fluorapatite was analyzed by Robinson et al. (1992) and the purple Fluorapatite was found to be slightly manganoan (0.2 wt % MnO). The purple Fluorapatite is not generally found in the complex phosphate parageneses; it tends to occur with Siderite and Brazilianite (Image 12).

Image 12: Purple Fluorapatite with Siderite and Brazilianite.

Image 13: Fluorapatite in-situ (Tait pers com; 2019)

Figure 8: Fluorapatite veinlet near the Whitlockite locality on the Big Fish River. The veinlet is cutting up-section at approximately right angles to the shale layers. No tectonism or altered layers are present at the vein margins. (Tait pers com; 2019)

Fluorapatite is most commonly short prismatic and colourless. Image 12 is an example of such a complex paragenesis with colourless Fluorapatite. One side of the vein is visible in outcrop at the edge of the cliff. (Image 13) Deep blue Lazulite is also part of the parageneses at this locality. The Fluorapatite veinlets (Figure 8) near the Whitlockite outcrop can also be monomineralic and produce tabular Fluorapatite crystals such as Image 14. There are a number of parageneses in the Rapid Creek locality that have the colourless Fluorapatite as a minor mineral associated with other phosphate phases; an example being the Goyazite-Fluorapatite-Wardite paragenesis.

Image 14: Tabular Fluorapatite from the veinlet in Figure 8: photo from Tomes (2013)


The primary occurrence of Gordonite is in the Variscite noduals of Clay Canyon, Utah (Wilson 2010). Here it occurs as late-stage cavity lining crystals in Variscite noduals altered to Crandallite. The Rapid Creek paragenesis of Gordonite is as a late-stage phase associated with Arrojadite crystals. There has been no new chemical data beyond that published in Robinson et al. (1992). The RRUFF database does not contain a sample of Gordonite from Rapid Creek.


Gormanite from Rapid Creek and the Big Fish River has more chemical data from Le Bail et al. (2003) and the sample in the RRUFF database. The landslide locality from Area C Locality 16 was noted in Robinson et al. (1992) but was written in more detail by Rod Tyson (written com. 1985); The Gormanite Locality 22 from E-Minerals is not found elsewhere in the literature.
The RRUFF Gormanite: 1 average sample posted with light and dark backscatter phases analyzed; n=6.
(Fe2+1.85Mg1.09Mn0.06)Σ=3(Al3.76Fe3+0.24)Σ=4(P1.00O4)4((OH5.65F0.35)Σ=6·2H2O ; (light phase) : (Fe2+1.53Mg1.42Mn0.05)Σ=3(Al3.89Fe3+0.11)Σ=4(P1.00O4)4((OH5.60F0.40)Σ=6·2H2O (dark phase)
Gormanite does not appear to vary in chemistry within the various sub-localities of the Rapid Creek area. Souzalite will occur only as zones within the Gormanite crystals, as it is approached by some of the individual analysis that are used to average the chemistry of the dark phase in the RRUFF sample.


The paragenesis is not mentioned in Robinson et al. (1992); see under Alunite Group for Goyazite and Wardite for chemical analyses of the individual phases. There are other parageneses that contain only Wardite-Goyazite and Wardite-Fluorapatite posted in the literature but they have no detailed sub-localities with the samples. The overlap of different parageneses from the literature can only be resolved by detailed mapping of the entire locality as the individual sub-localities of Robinson et al. (1992) are only locally applicable.

This is a fairly common paragenesis, as noted by numerous samples in the E-minerals database. This is one of the few parageneses that have a consistent mineralogy in the literature samples. It has identical Fluorapatite crystals in all photographed samples.


There is fairly extensive chemical data on this series in Robinson et al. (1992). The locality was described by Tomes (2013) who noted the different morphologies of Kryzhanovskite (pseudoctahedral) and Garyansellite (platy with bronzy schiller) without publishing the relevant chemical data. They say the Kryzhanovskite-Garyansellite veinlets are very thin and contain few if any other phosphate phases. Unlike the intricately intergrown Gormanite-Souzalite and Kulanite-Penikisite series the Kryzhanovskite and Garyansellite phases appear to form separate, macroscopically distinguishable crystals.
The RRUFF database Kryzhanovskite-Garyansellite: 1 average sample analysis with no significant zonation; n=10
(Fe3+1.69Mg1.18Mn0.09)(PO4)2(OH) 1.61·1.35H2O

The average sample is approximately midway between end member Kryzhanovskite and end-member Garyansellite. The crystals analyzed (Image 10) are pseudoctahedral with only minor evidence of schiller and have insignificant zonation in the backscatter image (Image 11). Most other occurrences of Kryzhanovskite (Moore, 1971) are the result of the alteration of primary pegmatite phosphates; they have major Fe and Mn and almost no Mg. The presence of a significant Garyansellite component in the Rapid Creek paragenesis suggests these are not the dehydrated equivalent of the Phosphoferrite Group minerals.

Image 15: Kryzhanovskite-Garyansellite analyzed sample; RRUFF database (2019).

Image 16: Backscatter image of RRUFF database (2019) analyzed Kryzhanovskite

In some cases samples have co-existing pseudoctahedral Kryzhanovskite without schiller and platy Garyansellite with strong schiller

Image 17: Kryzhanovskite (non-schiller on right) and Garyansellite (schiller on left) coexisting.

The Kryzhanovskite-Garyansellite co-existing sample is rare. The RRUFF database sample and the non-schiller pseudoctahedral crystals of Kryzhanovskite are more typical samples illustrated in the commercial literature. A M.Sc. thesis (Chu 2013) has been completed on the Reddingite Group at Rapid Creek (including Kryzhanovskite and Garyansellite). She used a variety of testing techniques to quantify the position in the Garyansellite-Kryzhanovskite-Landesite compositional triangle. It appears that at Rapid Creek Phosphoferrite is unstable and impure Kryzhanovskite is a stable hydrous Fe phosphate phase.


The Kulanite –Penikisite series was described with fairly extensive data by Robinson et al. (1992). There have been some notable finds since 1992. Kulanite +/- Penikisite is a common minor phosphate with many of the parageneses in both Rapid Creek and the Big Fish River phosphate suites. The crystals of Kulanite are variable in appearance but are not distinguishable from Penikisite visually (Mandrino et al., 1977). They range from thick, multi-faceted, twinned crystals that occur with Siderite such as:

Image 18: Thick bladed Kulanite; Stoneman Camp

To thin, often warped, plates associated with other phosphate phases.

Image 19: Thin bladed Kulanite: David Joyce photograph; Mindat.org.

RRUFF Kulanite: 1 average sample analysis with no significant zonation; n=10
(Ba0.97Sr0.03) Σ=1(Fe2+1.06Mn0.54Mg0.35Ca0.03Ti0.01 0.01) Σ=2(Al1.67Fe3+0.33) Σ=2(P1.00O4)3((OH) 2.68F0.32) Σ=3; OH estimated by difference and charge balance. The standard deviation within the individual analyses are not sufficient to make the sample not Kulanite but one analysis would bring it closer to the midpoint of the compositional triangle with (Fe0.98, Mn 0.58 Mg 0.39) 1.95.

Image 20: Analyzed Kulanite from RRUFF database (2019) associated with Fluorapatite.

There is insufficient data to determine the correspondence of the thinness of the Kulanite plates to changes in mineral chemistry. The grey scale on the backscatter image of the analyzed Kulanite does not vary in colour across the microprobe sample and the ten individual analyses do not vary from the average. Even though this RRUFF sample has a higher Bjarebyite component (Mn) than a Penikisite component (Mg) there have been no published reports of an analyzed Bjarebyite occurring in this paragenesis. Robinson et al. (1992) suggests that the chemistry of Kulanite varies with sub-locality at Rapid Creek and the thicker Kulanite such as Image 18 are close to end member Kulanite. There is no sub-locality for the RRUFF analyzed sample.


Robinson et al. (1992) provides extensive data on Lazulite especially for crystallography. Lazulite is one of the most common of the phosphate phases at Rapid Creek, and certainly the most famous. Many photographs of Lazulite occur in Mindat.org and within the commercial mineral literature. All Lazulite is non-end member, as the RRUFF database analyses illustrate: RRUFF Lazulite: 1 average sample analysis with no significant zonation; n=12 (Mg0.84Fe0.16)Al2.00(P1.00O4)2(OH)2. Lazulite is normally Mg rich and is often the only Mg dominant phosphate when it is in equilibrium with Fe dominant phases such as Gormanite and Kulanite.

Image 21: RRUFF analyzed Lazulite; Rapid Creek; (RRUFF database; 2019)

The photograph of the analyzed sample is typical of the Lazulite from Rapid Creek. The majority of the Lazulite samples are still found in the area of the original find at Cross-cut Creek. An expedition to the area by Kim Tait and team in 2012 recovered Lazulite samples in-situ (Figure 17) and illustrate the size and orientation of the Lazurite veins with the bedding of the shale horizontal (Figure 18).

The Lazulite veins vary from being open space filling veins with little tectonism on the edges of the vein and no significant displacement of the shale beds to veins with marked tectonism on the edges (Figure 19) and what appears to be shearing and local movement of the shale beds.

In-situ Lazulite with phosphatic shale

Rapid Creek, Dawson mining district, Yukon, Canada
Image 22: In-situ sample of Lazulite; Rapid Creek 2012; Tait (pers com.; 2019)

Image 23: Lazulite vein with sharp vein boundaries; Tait (pers com.; 2019)

Lazulite vein with shapr vein boundaries

Rapid Creek, Dawson mining district, Yukon, Canada
Image 24: Lazulite vein with tectonism at the edges. (Tait; pers. com 2019)

The rounded edges and non-planar faces on the Lazulite crystals that caused such problems for twinning and Miller indices assignment in Robinson et al. (1992) are still found in the modern Lazulite examples. A photo from Mindat.org illustrates the rounded edges and warped faces:

Image 25: Lazulite crystal with warped faces; Nadya Georgieva photograph; Mindat.org.


There is not much new analytical data for Ludlamite since Robinson et al. (1992).There is no RRUFF sample of or data for Ludlamite from the Rapid Creek occurrence.

The best Ludlamite-Vivianite locality is in Area C Locality 15, as mentioned in Robinson et al. (1992). A breccia sample from the locality has Ludlamite on one face and Vivianite concentrated on another.

Image 26: Ludlamite crystals

Image 27: Vivianite crystals from the same sample.

The images of Ludlamite-Vivianite in the Mineralogical Record Special Issue (Figures 55 and 57; Robinson et al. (1992)) have transparent blue Vivianite with the green Ludlamite crystals. However most of the samples from this locality have coal-black Vivianite that was never blue even when first recovered (Rod Tyson in an October 27, 2015 e-mail says that he recognizes the pictured sample in Image 26 and 27 and he collected it in 2005).


Alfredo Petrov (2019 Mindat.org website) has found Matioliite with Siderite, Wardite and Childrenite-Eosphorite from Rapid Creek. He confirmed, August 23, 2019, that it was a non-descript phosphate sample with a bit of Wardite that he sent to Dr. Tony Kampf for analysis and Matioliite is now confirmed from Big Fish River.


Robinson et al. (1992) has much data on Messelite but there has not been much published since. There is no RRUFF sample or data on Messelite from Rapid Creek.


Mindat has it in the Arrojadite-Fluorapatite-Metavauxite paragenesis with a 2018 EDS trace from Pavel Kartashov. The sample of Metavauxite looks very much like elongate Childrenite crystals and there may be more Metavauxite samples awaiting analyses. Pavel Kartashov confirmed, August 23, 2019, that it was found on a non-descript phosphate sample purchased for another phosphate without detailed sub-locality information. There is no RRUFF sample or data of Metavauxite from Rapid Creek.


The only reference in Mindat for Metavivianite is from Robinson et al. (1992). There is no RRUFF Metavivianite sample or data.


Tomes et al. (2013) have new data under her heading of Dypingite. The RRUFF database has one Rapidcreekite sample but has posted no analysis. The secondary minerals on joints in the Rapid Creek Formation form in coatings that may include several phases. Such a coating is Image 28.

Image 28: Secondary minerals that may include Rapidcreekite in situ (Tait pers comm.; 2019)


Satterlyite is a common mineral in the Rapid Creek Formation. It was mainly noted as a nodual mineral in Robinson et al. (1992) with a note on some of the non-nodular locations. Satterlyite has subsequently been found to be one of the major rock-forming minerals in the Rapid Creek Formation with many instances of bedding plane veins (Image 24) and non-nodular spray-like aggregates (Image 25) within the shale units (Tomes; 2013). Satterlyite has not been found in any of the vein deposits, though most of the veins cut Satterlyite-bearing shales.

Image 29: Analyzed Satterlyite sample; RRUFF database

RRUFF database has Satterlyite: 1 average sample analysis; n=15
(Fe2+7.95Mg2.25Fe3+0.66Na0.66Mn0.48)Σ=12(P1.00O3OH)(P1.00O4)5((OH)5.00O1.00)Σ=6; OH estimated by charge balance
The characterization of Satterlyite has a broad range of substitutions for Fe+2 and the backscatter image for the analyzed sample has no variability in the grey scale so these analyses represent an equilibrium composition. As it is found nowhere else the conditions for forming the multi-tonne quantities of Satterlyite within the Rapid Creek Formation must be unusual. The characterization of Nahpoite (Coleman and Robertson (1981) says: A brief study of compositions of nodules observed in situ suggests that 39% are composed primarily of pyrite, 58% of wolfeite, 2% of satterlyite and l% of maricite. An additional locality 5 kms north has similar nodules. Thus even in the noduals Satterlyite is not notably common. The noduals appear to be in equilibrium with the surrounding phosphatic shale, as outlined earlier, thus the crystallization of the Satterlyite must have been simultaneous within the shale and the noduals or there would be disequilibrium textures and zonation in the nodular Satterlyite.

Image 30: Satterlyite in-situ (Tait pers comm; 2019)

Image 31: In-situ Satterlyite spray in-situ with disruption of the sedimentary layers. (Tait pers comm; 2019)
The bedding plane vein and radial aggregates may grade into noduals but there does not appear to be the multi-minerallic composition of the bedding plane veins (the veins are either Satterlyite or Wolfeite) that is common in many of the noduals.


Not much more data has been published since Robinson et al. (1992). No Siderite samples have been posted on the RRUFF database.


Described in detail in Robinson et al. (1992) there have been many more finds since 1992. Wardite is another fairly common phosphate in the Rapid Creek paragenesis. It is described in a special issue of the Mineralogical Record on the “Yukon Phosphates” v. 23 no. 4. The Special Issue has a photo of the Wardite collecting locality at Area B Locality 12 with a person standing next to the creek side cliffs.

The Mindat.org photo gallery seems to imply that there are two distinct Wardite locations from here. The simple Wardite paragenesis is the most common with the multi-phase paragenesis (such as the Wardite with Goyazite and Fluorapatite) being discovered after 1992.

Image 32: Wardite on shale breccia vein

The Special Issue paragenesis diagram illustrates that this sub-locality is species poor with only Wardite and Eosphorite present. Tomes (2013) commented on this locality as it has no Lazulite in it; which is a ubiquitous phase everywhere else at RC-BFR. The Special Issue also notes the absence of Quartz and Lazulite with Wardite and suggests that it may be due to lower temperature crystallization. This would not work with the multi-phase Wardite localities that have Goyazite, Apatite, Quartz and Lazulite present.

RRUFF database Wardite: one sample with an average analysis: n=1
(Na0.93 0.05Ca0.02)Σ=1(Al2.97Fe3+0.03)Σ=3(P1.00O4)2((OH)3.88F0.12)Σ=4·2H2O


Robinson et al. (19921) has considerable data on Whiteite-(Ca, Fe,Mg) as it was characterized in 1978. The data of Capitelli et al. (2011) presents more data and determines the crystal structure of Whiteite-(Ca,Fe,Mg) and how it fits into the Jahnsite Group. The RRUFF database Whiteite-(Ca,Fe,Mg) has one sample noted by XRD but no published analysis. Most modern Whiteite-(Ca,Fe,Mg) samples occur as single crystals but the matrix samples occur on crystallized Quartz.

Image 33: Single Whiteite crystals on a Quartz crystal coated plate.


Vivianite from Big Fish River, Locality 15 Area C, has been noted in Robinson et al. (1992). The article has detailed descriptions of each of the phosphate minerals allowing most to be placed in their outcrop localities. Vivianite is a common mineral in phosphatic sediments, phosphate-bearing pegmatites and late-stage vein deposits where mobile iron and phosphorous are present in reducing conditions (Rothe 2016). Rapid Creek is somewhat unusual because there is abundant mobile Ca associated with Vivianite yet there is no recorded occurrence of Anapaite at Rapid Creek.

Detailed chemistry of most of the minerals, such as the Vivianite-Baricite group, is also provided. Rod Tyson, who collected this sample, has a photograph of the Big Fish River locality and some of the samples he collected from it, on his website.
A sample from the Tyson find is:

Image 34: Vivianite and Quartz

This sample is from a large boulder caught in one of the local landslides. Rod collected some of the best Vivianite and Gormanite from the boulder, which appears to have shielded the samples from the effects of the weather. The RRUFF sample has only the Big Fish River designation so it is impossible to get much closer, even though the samples appear to be similar. Arrojadite, Ludlamite and Vivianite are a characteristic assemblage for Locality 15 Area C and no other part of the Big Fish River has this suite of phosphates.

The pictured sample is a 4 cm vug in phosphatic ironstone with several 1 cm Vivianite crystals, brown spear shaped Arrojadite (K,Na), colourless Quartz and massive light green Ludlamite. A second sample from the same locality is listed under Ludlamite. It has similar Vivianite crystals that are coal black in colour.

The Dakota Matrix website has a photograph similar to the sample, as does Mineralatlas and many other mineral websites. The Vivianite crystals from this locality are distinct from the sedimentary Vivianite and are part of a series toward the magnesium analogue, Baricite. Each crystal is complexly zone from (Fe2.86, Mg0.06) to (Fe1.77, Mg1.03). The Baricite crystals and Vivianite crystals are physically indistinguishable. Baricite is unique to the Big Fish River area and does not occur in any of the other Vivianite localities.

The RRUFF database has a very similar sample from the Big Fish River location, looking more like the Ludlamite-Vivianite sample as it lacks Quartz. It was originally labeled as Metavivianite associated with Arrojadite, which occurs on both these samples. The microprobe polished section of the sample illustrated a Vivianite/Baricite intergrowth overgrowing cores of Maricite. The chemistry is: dark phase: (Fe2+1.77Mg1.21Mn0.02)Σ=3(P1.00O4)2·8H2O; H2O estimated by difference; light phase: Na1.00Fe2+1.00P1.00O4 (maricite). These are complex crystals with the three phases appearing to be in contact with one another with no obvious signs of reaction or corrosion. The “dark phase” is not uniformly dark but has zones of darker and lighter phases, possibly Vivianite and Baricite respectively. There are cleavable blue masses on the Ludlamite sample that match the photo very closely.

The RRUFF database Vivianite 1 average sample analyses co-existing with maricite; n=4
(Fe2+1.77Mg1.21Mn0.02)Σ=3(P1.00O4)2·8H2O; (dark phase, H2O is estimated by difference) : Na1.00Fe2+1.00P1.00O4 (light phase = maricite, with trace amounts of Mn)

Image 35: Analyzed Vivianite and co-existing Maricite from Big Fish River; RRUFF database

Image 36: Vivianite (deep blue) coatings in-situ; possibly with Baricite or Maricite (olive green plates included with Vivianite) (Tait pers comm; 2019)

Nodule Minerals

The RRUFF database has a number of analyses of the phases that occur as nodules in the Big Fish River area. They are:


One sample from the nodule locality has an average sample analysis; n=10
(Na1.80K0.20)Σ=2Fe2+1.00(Ca0.90Na2.10)Σ=3(Fe2+10.71Mg1.40Mn0.89)Σ=13(Al0.90Ti0.05Fe3+0.05)Σ=1(P1.00O4)11(P1.00O3OH)((OH)1.15F0.85)Σ=2; (OH) by stoichiometry and charge balance


One average sample; n=8
(Mg1.66Fe1.33Mn0.01)Σ=3(P1.00O4)2·8H2O; H2O estimated by difference


(nodule with maricite) One average sample analysis; n=8


One average sample analysis; n=10


One average sample analysis; n=15
(Fe2+7.95Mg2.25Fe3+0.66Na0.66Mn0.48)Σ=12(P1.00O3OH)(P1.00O4)5((OH)5.00O1.00)Σ=6; OH estimated by charge balance


A RRUFF database sample is associated with Satterlyite in a nodule but has no chemical analysis. A subsequent article to Robinson et al. (1992) is on the crystal structure of Wicksite (Cooper and Hawthorne (1997)). They say that Wicksite is structurally identical to Grischunite (Na Ca2Mn4 (Mn, Fe) (AsO4)6.2H2O)


One average sample analysis; n=10
(Fe2+1.72Mg0.22Mn0.06)Σ=2P1.00O4((OH)0.87F0.13)Σ=1; (OH) estimated by difference and charge balance
Kolitsch (2004) says: Mg‐rich wolfeite [diiron(II) hydro¬xide phosphate], (FeII,Mg)2(PO4)(OH), from the Big Fish River area, Yukon Territory, Canada, is isotypic with its MnII‐dominant analogue triploidite.

The mineral phases from the nodules differ from the vein parageneses because almost all of the nodular minerals are iron and/or magnesium dominant. They often occur with Pyrite whereas Siderite is the common non-phosphate phase in the veins. There does not appear to be replacement of one phosphate mineral by another though the relationship between the phosphate and pyrite nodules is complex and not determined as yet.

The mineralogy of the phosphate nodules does not vary significantly within the exposures along the Big Fish River (Tomes 2013). The variation within the nodule mineralogy appears to be vertical, from the base of the Rapid Creek Formation toward the top of the stratigraphic section. The ratio of phosphate: pyrite nodules increases toward the top of the section so any replacement of phosphate by pyrite is regional and is not a response to local fS2 changes. There does not appear to be zonation in the mineralogy of the individual nodules and their distribution within the stratigraphic section is independent of the mineralogy of the individual nodules. Nodules similar to those in the Rapid Creek Formation are present in other phosphorite sequences (Dar et al. 2017).

The iron content of the Rapid Creek Formation is higher than average for worldwide phosphorites which are normally found in iron-poor environments and are generally interbedded with shales, cherts, limestones, dolomites and, more rarely, sandstones (Dar et al. 2017). The pH/eH parameters that are noted from the other worldwide phosphorite horizons may not be applicable to the Rapid Creek phosphorite as Satterlyite appears to be the primary phosphate phase at Rapid Creek, even though it was not noted as an unknown mineral in the bulk analyses of Brock (1975). The high-iron content is reflected in the mineralogy of the Rapid Creek nodules (Satterlyite, Wolfeite, Wicksite etc.); normally these nodules in a phosphorite are carbonate-apatite, sometimes pseudomorphed after fossils.

The iron content at Rapid Creek is not a function of its derivation from the Pre-Cretaceous strata of the Richardson Mountains as they are not particularly iron rich. Most of the clastic wedges of mid-late Cretaceous age in the eastern Cordillera are notably iron-poor. There are few occurrences of Pyrite within the Jurassic to Cretaceous Age coal measures in British Columbia and Alberta. The cobble to boulder sized clasts within the co-existing conglomeratic facies are iron-free chert. Even the deeper water phosphatic shales of the equivalent formations in the Williston Basin are not enriched in iron. Their phosphate mineralogy is standard carbonate apatite of the “Fish Scale” horizons.

The structures within the phosphatic shales of the Rapid Creek Formation match many of the structures in other phosphatic sequences but the bulk chemistry is very different. It is possible that the iron content of the eastern Cordilleran clastic wedges was mobilized during the deposition of these wedges in a tropical to sub-tropical coastal plain setting. The resulting mobile iron was deposited at the northern end of the seaway, in the Rapid Creek Formation, due to changes in temperature or pH of the seaway water at the interface with the open ocean. It would convert normal phosphatic shale to the unique iron-phosphate bearing strata of the Rapid Creek Formation.

The pressure and temperature of the vein-type phosphates was determined to be 180oC to 200oC by Robinson et al (1992) but the pressure and temperature of the co-existing nodules was not attempted, even though quartz segregations, contemporaneous with the existing phosphates, exist within the nodules. The nodules cannot have formed at a higher temperature than the veins as there are no alteration halo around the nodules, as there sometimes is around the veins.

Secondary Minerals

Not much done since Robinson et al. (1992) beyond the RRUFF sample of Rapidcreekite. Tomes (2013) has a paragraph on the secondary phases from Rapid Creek:

Dypingite Mg5(CO3)4(OH)2·5H2O, closely associated with hydromagnesite Mg5(CO3)4(OH)2·4H2O and lansfordite, was collected from an open fracture system above permafrost in the Rapid Creek — Big Fish River area in August 2012. Dypingite occurs in a botryoidal habit consisting of very fine radiating crystals forming a translucent crust of spheroidal aggregates. Some of the spheroids can be up to 1 mm in diameter. The fracture system is on a northwest facing slope and dypingite is found 10 cm or more below the surface coating surfaces on open fractures. The dypingite is often covered with loose and chalky hydromagnesite that is brittle and falls away from the fracture surface. The relationship suggests that the hydromagnesite forms as a dehydration product of dypingite at this locality. The Big Fish River occurrence of dypingite is unusual as it is not related to the alteration of ultramafic rocks but is located within a sequence of phosphate-rich sedimentary rocks. However, a magnesium sulfate, epsomite (MgSO4·7H2O), is commonly coating cliff faces throughout the area.

Many of the secondary minerals are water soluble or decomposable in water and so occur only on sheltered rock faces. Image 28 shows a typical fracture face on a block of shale with white secondary phases coating one face. The weathered crusts on the shale are widespread within the Rapid Creek Formation and almost every protected face is coated in a fine-grained white encrustation (Figure 9).

Figure 9: Rapid Creek Formation outcrop coated with white secondary phases (Tait pers. com, 2018)

Magnesium, but not the more abundant iron, is the major mobile cation in these secondary mineral phases. Iron oxide is present as Goethite, but it is a minor mineral within the secondary phases (Robinson et al., 1992). There is no evidence for widespread oxidation of the Siderite component of the shales as the Goethite occurs only on small fracture zones and the Siderite crystals within the secondary phosphate veinlets are normally unaltered.

There is minor mobilization of silica during veinlet formation as David Joyce has noted two generations of Quartz within the phosphate veinlets. Chalcedony coatings have not been noted in the magnesium carbonate coated joints and fractures so there is no active deposition of silica. There have been no studies of the edges of the phosphate veinlets to categorize the changes in mineralogy across the veinlets and if there is depletion in phosphate content of the bulk shale in the immediate 5 to 10 centimeters of the veinlet edges. Studies by Armbruster et al. (1996) and Heijboer (2006) on the Swiss alpine clefts reveal changes in the surrounding rock with precipitation of the cleft minerals from rocks adjacent to the clefts. It is possible the secondary phosphate veinlets at Rapid Creek are similar to the Alpine-style clefts as these clefts have been found in many countries besides Switzerland, always in association with mountain building metasediments such as the Rapid Creek Formation. The cleft minerals are formed from the chemical ingredients in the adjacent strata and it appears as if the Rapid Creek vein parageneses form in the same manner.

Questionable Minerals

Some questionable minerals have been reported in the literature for which there is insufficient available data to categorize them. They are:

1): Unknowns in the 1992 Special Issue; some may still be new. There are currently no unknown minerals listed in the Mindat.org locality page for Rapid Creek/Big Fish River but the six unknown from Robinson et al. (1992) are still listed as unknowns on the Dawson Mining District page with Robinson as the only reference.
2): Xanthoxenite with Arrojadite R. Mielke analyses #404 and #403. I contacted Reiner, on August 22, 2019 for further information on their new finds. Reiner has not responded.
3): Variscite: E-minerals have Variscite in this paragenesis.
4): Scholzite: E-minerals have Scholzite with Collinsite in this paragenesis

Phosphate Mineral Geochemistry:

The differences between the phosphate parageneses at Rapid Creek-Big Fish River and alteration sequences within the pegmatite phosphates can be summarized as:

1): Temperature of formation;
2): Availability of iron;
3): Absence of lithium and boron;
4): Non-end member composition of the phosphate phases;
5): Non-localization of the individual parageneses;
6): Primary phosphate mineralogy;
7): Absence of common minerals in the parageneses

1) The temperature of formation of the Rapid Creek paragenesis has been investigated by Robinson et. al (1992) at 180o-200o C. An unresolved question in that article was to correspondence of many of the Rapid Creek minerals with corresponding pegmatitic phosphate phases that have been noted to form at very much higher temperatures than is possible at Rapid Creek (Moore (1973). Subsequent data from other phosphate occurrences that formed at low temperatures indicate that the pegmatite phosphate sequences are distinct from the sedimentary phosphate parageneses at Rapid Creek and elsewhere and the lower temperatures of formation are characteristic of the phosphates found in phosphatic shales.

2): A suite of low-temperature iron phosphates, including fluorine dominant Fluellite, has been found in brecciated fluorapatite-containing phosphatic magnetite in the Leveaniemi Mine, Sweden (Bjallerud, 1989). These phosphates are transitional to the limonite phosphate suite and contain Kidwellite-Laubmannite-Strengite-Rockbridgeite-Varscite-Wavellite similar to the Indian Mountain, Alabama and Dug Hill, Arkansas occurrences. Having Fe and Al dominant phosphates occurring in equilibrium seem to be characteristic of this type of deposit, whether the Al phosphates are Wavellite-Variscite (Dug Hill) or Augelite (Rapid Creek).

The Al dominant phosphate-bearing sequences in the Arkansas Novaculite at Dug Hill, Arkansas (Fluorwavellite and Variscite; Mindat.org) and in the non-base metal enriched Carlin-type gold deposits in Nevada such as the Willard Mine (Fluellite, Fluorwavellite, Variscite and Minyulite; Jensen and Leising (2001)) are intermediate between the pegmatite phosphates and the Rapid Creek phosphatic shales. The more complex Carlin –type gold deposits such as the Gold Quarry Mine, Nevada (Jensen, Rota and Foord (1995)) and The Candelaria District, Nevada (Adams (2019)) have simple hydrous and fluorine-bearing phosphates as well but have a much larger overprint of Cu-Zn-Cd arsenates and phosphates in addition; such as Goldquarryite (Roberts et al. (2003)).

In the simple cases fluorine-bearing secondary phosphate minerals occur in late-stage cavities in tectonized phosphatic shale and novaculite. In the phosphate zone at the Willard Mine goethite is common; calcite is rare and siderite is not present (Jensen and Leising (2001). The Arkansas novaculite has minimal Goethite and no Calcite or Siderite.

Samples of these parageneses are illustrated as:

Image 37: Fluorwavellite hemispheres in Novaculite

Image 38: Fluellite and Fluorwavellite in a cavity in Phosphatic Shale; Willard Mine

Fluorwavellite, Variscite, Minyulite and Fluellite do not occur in the Rapid Creek paragenesis. This may be due to the lack of fluorine in the Rapid Creek system. There have been many analyses to check for F in the hydrous Rapid Creek phosphates. Robertson (1980) assumed there was no fluorine and gave his analyses as 100% OH/H2O. Robinson et al. (1992) mentions a few phases with a minor fluorine content but none fluorine dominant. The RRUFF database has several phosphate phases with minor fluorine but only “Bobdownsite” with dominant fluorine in the appropriate crystal site; this result has been disputed (McCubbin (2018)).

The lack of common phosphates in the Rapid Creek parageneses may be due to other unknown factors because the Willard Mine and Dug Hill phosphate suites are fairly common worldwide in suitable geochemical settings and the Rapid Creek phosphate parageneses are unique so far.

In contrast to the Rapid Creek and Willard Mine phosphatic shales the Permian Age Meade Peak Member of the Phosphoria Formation is a major resource for phosphate rock. It occurs over an area of 11, 600 square kilometers in western Wyoming, southeastern Idaho and northeastern Utah and it has been extensively studied and the mineralogy of its weathered and non-weathered portions delineated. The Meade Creek phosphorite is richer in P2O5 than the Rapid Creek Formation with: 80% apatite; 10% quartz; 5% muscovite-illite; 2% organic matter; 1% dolomite-calcite; 1% iron oxide and 1% other components (Derkey and Palmer (1976)). While there have been some secondary vanadium and phospho-vanadium minerals described (Phosphovanadylite-Ca; Medrano (1998) and Sincosite) there have been no significant parageneses of secondary phosphates found in the Meade Creek Member.

3): Lithium and boron bearing minerals, besides accessory detrital tourmaline in some of the clastic sediments, do not occur in any of the phosphatic shale sequences. The secondary pegmatite phases such as Lithiophosphate that occur in cavities in the Li-bearing Tanco Mine (Gunter 2018) would not be possible in any of the phosphatic shale sequences. Many of the secondary phosphates that are found in the late stage cavities in pegmatites worldwide are lithium-bearing because of the chemistry of the major primary pegmatite phosphate: Triphylite-Lithiophyllite. Occasionally non-lithium bearing primary phosphate minerals occur, such as in the Cross Lake pegmatite suite (Gunter 2018). Here Bobfergusonite, Manitobaite and Beusite are the main primary phosphates. There are no known secondary phosphates at the occurrence.

4): One of the primary differences between the Rapid Creek hydrous phosphates and the other types of hydrous phosphates is the variability of the Fe:Mn:Mg ratios in the Rapid Creek parageneses and the lack of end member compositions. Most iron/manganese/magnesium phosphate occurrences are near end member. An example of a simple iron phosphate/arsenate/sulphate occurrence that has been chemically analyzed, the Pharmacosiderite-Dufrenite-Jarosite occurrence at the Lone Tree Mine, Nevada (Dyar et al. 2018) is virtually end member Dufrenite and Pharmacosiderite associated with end member Jarosite in the oxidation zone of a Carlin-type gold deposit. The secondary phosphate, arsenate and sulphate minerals occur as euhedral crystals lining cavities in fine-grained silica.

Image 39: Pharmacosiderite, Dufrenite and Jarosite on silica: Lone Tree Mine, Nevada

In this case the anions rather than the cations were mobilized and the three phosphate-arsenate-sulphate mineral phases occur as co-existing crystals without evidence of disequilibrium.

Barium and strontium-dominant phosphates are unusual outside of the Rapid Creek parageneses. A detailed article on the secondary phases in the complex Carlin-type Candelaria District, Nevada (Adams, 2019) including SEM, EDS, XRD, RS and FTIR investigations into the chemistry of the phases, has done much to sort out the compositions of the various mineral series that occur in this type of environment. The Mg and Fe dominant phosphates: Cacoxenite, Collinsite, Gordonite, Mitridatite, Montgomeryite, Overite, Strengite, Whiteite-(Ca,Mg.Mg)and Whitlockite occur with calcium/aluminum phosphates: Crandallite, Fluorapatite, Fluorwavellite and Variscite-Metavariscite.

Another complex Carlin-type deposit at the Gold Quarry Mine has: Mg-Fe phosphate Leucophosphite, Phosphofibrite, Strengite and Tinticite associated with Al-Ca phosphates: Augelite, Carbonate Fluorapatite, Crandallite, Englishite, Fluellite, Kingite, Variscite and Wavellite (Fluorwavellite?) (Jensen, Rota and Foord 1995). There are no Ba or Sr dominant phosphates in either the Candelaria or Gold Quarry parageneses; though Baryte is a common phase in the Gold Quarry Mine.

A fluorine rich Fe-Al phosphate occurrence within a sedimentary phosphorite formation is the Silver Coin Mine, Nevada (Adams, Wise and Kampf 2015)). It is a far smaller locality than the large mines in the Candelaria District or the large-scale Gold Quarry mine. The Phosphate Stope locality in the Silver Coin Mine is approximately 25 meters by 25 metres in size. It is an unusual locality with many co-existing individual members of: the Adelite Group, the Beudantite Group, the Plumbogummite Group, the Brackebuschite Group and the Strengite Group. Phosphate species include: simple hydrous phosphates and mixed hydrous sulphate-phosphates, hydrous arsenate-phosphates, hydrous vanadate-phosphates and hydrous silicate-phosphates. Fluorine-dominant phosphates include members of the Hedyphane Group, the Apatite Group and the Wardite Group, (including the new mineral Fluorowardite) and Fluorwavellite. Baryte and Fluorite are common minerals in the Phosphate Stope; often associated directly with the secondary phosphate minerals. A typical phosphate assemblage sequence is:

1) Baryte (Ba)
2) Smithsonite replaced by Goethite (Fe)
3) Lipscomite, Kidwellite (Fe phosphate sequence)
4) Fluorapatite, Fluorowardite, Turquoise, Variscite, Wavellite-Fluorwavellite (Al-F phosphate sequence)

No barium or strontium dominant phosphates exist at this locality; even though Ba is mobile under these conditions. One magnesium phosphate, Montgomeryite, occurs here but it is in the arsenate paragenesis. The variability of the phosphate mineralogy has been ascribed to a complex history of alteration with many co-existing micro-environments.
There are a series of iron phosphate occurrences in limonite deposits in and around the Appalachian Mountains from Alabama northward to New Jersey. These iron phosphates have a well-studied mineralogy (Barwood (1974) and Gordon and Hollabaugh (1989)) and their paragenesis is different than the western phosphate occurrences. They are dominated by Eleonorite, Strengite and Kidwellite; unusual minerals in the western phosphates. Kidwellite is the only phosphate in the western parageneses; being found at the Silver Coin Mine.

Unlike the phosphates minerals, the Fe-Mn-Pb arsenate minerals in many parageneses have significant variations and solid solution series. A regionally extensive arsenate sequence of: Mimetite-Alunite supergroup (Beudantite/Segnitite/Tsumcorite)-Carminite is comparable in occurrence to the Rapid Creek phosphate suite. The suite is found in several base metal deposits worldwide in an almost identical paragenesis. Detailed descriptions of this paragenesis come from: the Krupka Deposit, Czech Republic (Sejkora et al. (2009)); the Mohawk Mine, California (Wise (1990)) and the San Rafael Mine, Nevada (Custor and Ferdock (2004)). They indicate that while the colourless Mimetite and red Carminite are close to end member composition the Alunite supergroup phases in this paragenesis have a wide range of compositions. This is similar to the Rapid Creek occurrence where phases such as Gormanite and Kulanite may have a wide range of chemistry while the co-existing Brazilianite and Augelite has no significant chemical variability.

5) The four fold designation for the mineral parageneses of Robertson (1980) has fallen apart with the discovery of many more vein localities but there are still patterns of chemical distribution. The classification of parageneses at Rapid Creek is sometimes difficult due to the very general locality for many of the commercial samples; Rapid Creek or Big Fish River being generally all that is available. The phosphate nodual species are the only suite of phases for which the exact locality is known. In most other cases the parageneses can occur in multiple localities; as was listed in the locality compilation of Robinson et al. (1992). There are so many other occurrences that have been discovered since 1992 that many of the localities are now considered “areas” rather than individual veins (Tomes 2013). Only the original Gorceixite find from Area A, Locality 1 and the euhedral Collinsite from Area B Locality 11 are confined to one outcrop. The Whitlockite vein may be a similar one outcrop occurrence but there is not enough data yet to be certain.

Aside from the suite of individual mineral analyses from the RRUFF database a suite of minerals used for the Mossbauer Analyses (for cation assignment and crystal structure) by Dyar et al. (2014) includes 9 samples from the Rapid Creek/Big Fish River area. These samples are:
1): Bobierite =Vivianite group;
2): Gormanite;
3): 2 samples of Kulanite;
4): 2 samples of Lazulite;
5): Maricite;
6): Satterlyite and
7): Whiteite=Jahnsite group.
The structural analyses mention, though they did not analyze, phases from Rapid Creek/Big Fish River area of:
1): Kryzhanovskite and Garyansellite = Phosphoferrite group;
2): Baricite=Vivianite group and
3): Wolfeite=Triplite group.

These 9 analyzed samples from Rapid Creek/Big Fish River area are seldom end member compositions and the XRD analyses do not always conclude that the samples used for the Mossbauer analysis are 100% of the targeted mineral phases. Dyar et al. (2014) still considered the data sufficiently good that cation assignments of the targeted phases could be made.

6): The primary phosphate mineral in the Rapid Creek Formation and other phosphatic shales mentioned in the other occurrences is Fluorapatite-Hydroxyapatite. Only rarely is massive Fluorapatite found in pegmatites (Cross Lake Pegmatite # 36; Gunter 2018).and it does not corrode or participate in the creation of secondary phosphate phases; Fluorapatite is often created as a secondary mineral in pegmatite cavities. In the Rapid Creek Formation there appears to be an X-Ray amorphous phases or phases of iron phosphate (possibly equivalent to Santabarbaraite) that has never been investigated. Its chemistry is unknown but it may be a primary source of chemicals for the crystallization of the secondary phosphate phases.

7): The Rapid Creek Formation is noted not only for the presence of well crystallized secondary phosphate minerals but also for the absence of common phosphates that are a normal part of the diagenesis of iron-phosphate bearing sediments. Nriagu and Dell (1974) in their analyses of the diagenetic basic iron phosphates indicate that the most common hydrous iron-manganese phases formed during diagenesis should be Vivianite, Ludlamite, Reddingite (Phosphoferrite Group) and Anapaite. Vivianite and Ludlamite are constituents of the Rapid Creek secondary phosphate paragenesis and the Kryzhanovskite-Garyansellite Series is the member of the Phosphoferrite Group present here. Anapaite, though the Eh and pH overlap part of the Vivianite composition space, (Nriagu and Dell, 1974) and mobile Ca is present in almost all the Rapid Creek parageneses, has not been found.
Nriagu and Dell (1974) consider Strengite to be formed within an elevated Eh composition field (> 0.1) and thus it should not be a part of a diagenetic iron-phosphate suite. Strengite has not been found in the vein deposits at Rapid Creek; only as a very rare mineral in the noduals (Robinson et al., 1992). The Indian Mountain and Swedish Strengite occurrences are low-temperature deposition within limonite cavities and their presence was unexplained in Nriagu and Dell (1974). Woodhouseite and Svanbergite are two of the Alunite Group minerals that have the proper chemistry to occur in the Rapid Creek paragenesis and do occur in late-stage phosphate sequences at the Champion Mine, California and Mount Brusillof, British Columbia respectively. Stoffregen and Alpers (1987) investigate the mobility of phosphate ions under low-temperature hydrothermal alteration and note Woodhousite and Svanbergite in the upper levels of three hydrothermal coper-gold deposits.

Other common minerals in digenetically altered phosphatic sediments are Calcite, Dolomite, Wavellite and Variscite. These occur in almost all other digenetically altered sediments where mobile Ca, Mg, PO4 and CO3 are present. The presence in the Rapid Creek phosphates of Augelite rather than Variscite and/or Wavellite may be temperature dependent.

Wise and Loh (1976) in their study of phosphate-bearing Andalusite-Kyanite deposits synthesized Augelite, Lazulite and Scorzalite but not Variscite or Wavellite, at temperature runs from 395oC to 509oC. These temperatures are far higher than the 190oC of the Rapid Creek phosphate veins, found by fluid inclusions (Robinson et al. (1992). Minor Augelite does occur with abundant Wavellite, minor Variscite and the Vauxite Group phosphates in the late-stage paragenesis at Llallagua, Bolivia (Hyrsl and Petrov, 2006) but Augelite is so rare its relationship to the other phosphates was not recorded. Wardite with abundant Variscite and other phosphates (but not Wavellite) occurs at the very low temperature Clay Centre Variscite Mine, Utah (Wilson, 2010).

A Ludlamite-Vivianite paragenesis, similar to the Rapid Creek paragenesis, occurs in the late stage secondary and primary sulphide zones of the Blackbird Mine, Idaho (Young, 2010). These are fracture and vug filling phases in massive Pyrite in a cobalt mine. Ludlamite and Vivianite are also found as late-stage hydrothermal vug lining in other large, complex hydrothermal deposits such as Santa Eulalia, Mexico (Megaw, 2018) and Trepca, Kosovo (Feraud, Maliqi and Meha, 2007). Few other phosphates occur in these polymetallic deposit’s parageneses.

The Rapid Creek phosphorite has an unusual mineralogy but it is not unique in the Western Canadian Sedimentary Basin. Phosphatic shales, identified by the presence of “fish-scales” and “spotty shales” from abundant vertebrate remains and fish coprolites are common in many areas of the Jurassic and Cretaceous sedimentary sections in the Western Canadian Sedimentary Basin (Western Canadian Sedimentary Atlas, 1984). These shales are far to the east of the Cordilleran Disturbed Belt and so have not had any of the folding, faulting and vein formation of the Rapid Creek Formation; and thus none of the secondary phosphates. The phosphatic shales are almost always encountered in the subsurface so alteration of the phosphates from carbonate fluorapatite to Wavellite etc. might not be noticed.

The absence of clay minerals is one of the main differences between the Rapid Creek Formation and the other contemporaneous sedimentary sequence of the eastern Cordillera. Traces of volcanic ash in the form of thin bentonite layers are widespread in the shoreline to shallow marine mid-Cretaceous sedimentary sections as far east as Manitoba. These bentonite-sourced smectite clays are not present in the Rapid Creek Formation and provide supporting evidence for the deep-water deposition of the phosphatic shales. Results from Wise and Loh (1976) indicate that Al will partition into Al phosphates instead of Al silicates if there is an abundance of PO4 in the fluids. Al phosphates and Quartz are abundant in the Rapid Creek parageneses and clay minerals are absent in the vein and nodual portions of the Rapid Creek Formation and only Kaolinite is present in the bulk analysis of the phosphatic shale (Brock, 1975).

A suite of localities that have only mobile silica and Siderite, as does the Rapid Creek Formation, are the Columbia River basalts in the vicinity of Spokane, Washington and the Clackamas River valley, Estacada, Oregon. In these localities a particular member of the Columbia River basalt suite (the Grande Ronde Basalt Unit) is flooded with silica (Shannon 1923) and large amounts of Siderite, with only minor Mn and Mg, Pyrite and “iron opal” (now opal) are deposited in open cavities in otherwise unaltered basalt. Late-stage Calcite but not Dolomite has been described in these cavities; the plagioclase component of the basalt would provide the mobile Ca ions. A sample from the Clackamas River valley:

Image 40: Siderite and Opal on basalt, Clackamas River, Oregon

illustrates the cavity lining deposition of both the Opal and the Siderite. The mobility of the iron carbonate is limited to the formation of multi-crystalline spheres or hemispheres of Siderite. The basalt matrix of the sample is glassy and does not have the typical basaltic texture.

In other low-temperature phosphorites the cavity fillings are dominantly carbonate-apatite. One of the early sources of phosphorous for fertilizer was the off-shore “guano” islands in the Caribbean Sea, Indian Ocean and the South Pacific Ocean (Dar et al., 2017). On these islands the soluble phosphates from bird guano had infiltrated the underlying limestone and other bedrock to produce rock phosphate. This resource is not currently commercially available but it has produced a number of Ca and Mg phosphate minerals from low-temperature cavities in the bedrock.

The “bat guano” cave deposits, noted by Dar et al., (2017), also produce a different suite of phosphates as listed in Mindat.org. As these caves are not open to the atmosphere and ambient rainfall the phosphate species within them do not correspond to the phosphate suite in the Rapid Creek Formation. There is however a bi-phosphate species in the cave deposits, Brushite, which occurs as a rare mineral in the secondary phases at Rapid Creek in areas protected from rainwater.
The overall phosphate content of the Rapid Creek Formation is not significantly higher than that of the Phosphoria Formation so the variability in phosphate mineralogy is not related to the initial phosphate loading from the source area.


The Rapid Creek/ Big Fish River area in the northeastern Yukon Territory is a world-renown set of occurrences of rare to very rare phosphate species. Often these species occur in euhedral crystals that are the best in the world. The enclosing Rapid Creek Formation is an unusual phosphatic iron conglomerate-sandstone-shale sequence.

The Rapid Creek Formation is part of a series of Albian Age sediments that were deposited along the eastern edge of the rising Cordillera. Most of the Albian Age Cordilleran sedimentary packages were deposited in tropical to temperate settings and often contain coal measures (Gates Formation etc.). Storm deposits are common in the near-shore facies with chert-boulder conglomeratic units in the coal-bearing sequences and heavy-mineral enriched sands in the Burmis area of Alberta. The Rapid Creek Formation has been suggested by Yeo (1992) to be a sub-arctic deposition, formed northward of the other Canadian Cordilleran sedimentary packages. Sedimentary analysis of the formation indicates that it is turbidite-based deep water sequence. Its deep-water deposition means it does not contain the storm deposits of the shallower-deposited sedimentary sequences. The Rapid Creek Formation does have conglomeratic slump deposits; possibly formed in response to the large surface storms.

The majority of world-wide phosphatic shale sequences are shallow-water deposition, some close to the wave base. The Rapid Creek Formation is unusual both for its deep-water deposition and for its iron and phosphate rich chemistry. All other phosphorites and phosphatic shales have fluor/hydroxyl apatite as the primary phosphate phase, while it appears the primary phosphate phase in the Rapid Creek Formation is Satterlyite with minor Fluorapatite. This unique primary chemistry is also reflected in the unique parageneses of the late-stage phosphate minerals; which were derived from the immediately surrounding rocks in a similar style to the Alpine cleft occurrences world-wide.


I wish to thank Dr. Kim Tait and the staff of the Royal Ontario Museum for their assistance in providing the field photographs and Dr. Tait’s assistance with the theses material. The ROM photos are published with permission of the Royal Ontario Museum; Courtesy of the Royal Ontario Museum, copyright ROM.

Some of the mineral photographs are courtesy of Mindat.org and the RRUFF published databases. When not my photos the publisher of the photographs is mentioned and they are thanked for their contributions.


Adams P.M. (2019) The Candelaria District, Mineral County, Nevada: Mineralogical Record v. 50 no. 2 p. 125-168

Adams P.M., Wise W.S. and Kampf A.R. (2015) The Silver Coin Mine, Iron Point District, Humboldt County, Nevada: Mineralogical Record v.46 no 5 p. 701-734

Armbruster Th., Kohler Th. and Meisel Th. (1996) The zeolite, fluorite, quartz assemblage of the fissures of Gibelsback, Fiesch (Valais, Switzerland): crystal chemistry, REE patterns and genetic speculations: Schweizerische mineralogische und petrographisch Mitteilvegen; Band 76 Heft 2, P. 131-146

Bayliss P., Kolitsch U., Nickel E.H. and Pring A (2010) Alunite Supergroup: recommended nomenclature: Mineralogical Magazine v. 74 (5) p. 919-927

Brock J.S. (1975) Exploration Report; Delta Iron Deposit; NTS 117-A: Mining Recorder, Whitehorse Y.T. 9 p.

Buhmann D., Buhmann C. and von Brunn V. (1980) Glaciogenic banded phosphorites from Permian sedimentary rocks: Economic Geology v. 84 p. 741-750

Camara F., Oberti R. Chopin C.and Mendenback O. (2006) The Arrojadite enigma: 1. A new formula and a new model for the arrojadite structure: American Mineralogist v. 91 p. 1249-1259

Capitelli F., Chita G., Cavallo A., Ballatressa F. and della Ventura G. (2011) Crystal structure of Whiteite-(CaFeMg) from Crosscut Creek, Canada: Z. Kristallogr. V. 226 p. 731-738.

Chopin C., Oberti R and Camara F. (2006) The Arrojadite enigma: II. Composition space, new members and nomenclature of the group: American Mineralogist v. 91 p. 1260-1270

Chu K.W. (2013) The Phosphoferrite group: variations in chemical composition of minerals found in Rapid Creek, Yukon Territory, Canada: B.Sc. Thesis, University of Toronto, 35 p.

Cobic A., Zebec V., Scholz R., Bermanec V. and de Brito Barrento (2011) Crystal morphology and XRD peculiarities of brazilianite from different localities: Nat. Croat. V. 20 no. 1 p 1-18

Cook P.J. and McElhinny M.W. (1979) A reevaluation of the spatial and temporal distribution of sedimentary phosphate deposits in light of plate tectonics; Economic Geology v. 73, p. 339-350

Cooper M.A. and Hawthorne F.C. (1997) The Crystal Structure of Wicksite; Canadian Mineralogist, v. 35, pp. 777-784

Custor S.B. and Ferdock G.C. (2004) Minerals of Nevada, Nevada Bureau of Mines and Geology, Special Publication 31

Dar S.A., Khan K.F. and Birch W.D. (2017) Sedimentary Phosphates; Reference Module in Earth Systems and Environmental Science; Elsevier

Dekey P.D. and Palmer P. (1976) Maps showing selected geology and phosphate resources of the Meade Peak Quadrangle, Bear Lake and Caribou Counties, Idaho: USGS Map MR-76 notes. 6 p.

Dyar M.D., Jarwin E.R., Breve E., Marchand G., Nelms M., Lane M.D., Mertzman S.A. Bish D.L and Bishop J.L. (2014) Mossbauer parameters of iron in phosphate minerals: Implications for interpretation of martian data: American Mineralogist v. 99 p. 914-942

Dzikowski T.J. (2004) The Crystal Chemistry of Gorceixite, Grandidierite and Traskite; M.Sc. Thesis, University of British Columbia, 80 p.

Dzikowski T. J., Groat L.A. and Jambor J.J. (2006) The symmetry and crystal structure of gorceixite, BaAl3[PO3(O,OH)]2(OH)6, a member of the Alunite Supergroup: Canadian Mineralogist v. 44 p. 951-958

Feraud J., Maliqi G and Meha V. (2007) Famous Mineral Localities: The Trepca mine, Stari Trg, Kososvo: Mineralogical Record v. 38 no. 4 p. 267-302

Gordon J.B. Jr. and Hollabaugh C.L (1989) Phosphate microminerals of the Indian Mountain area: Mineralogical Record v. 20 no. 5, p. 355-362

Gunter W.R. (2018) Pegmatite Mineralogy: Cross Lake, Manitoba: Mindat.org Article.

Gunter W.R. (2018) Secondary Hydrothermal Minerals at the Tanco Mine: Mindat.org article.

Heijboer T.C. (2006) Origin and pathways of pro- and retro-grade fluids, PTt paths and fluid-mineral equilibria from Alpine veins of the Central Alps: Case studies of the Fibbia and Amsteg areas: PhD Thesis, University of Basel, 173 p.

Hyrsl J and Petrov A. (2006) Famous Mineral Localities: Llallagua, Bolivia: Mineralogical Record v. 37 no.2 p. 117-164

Jensen M. and Leising J. (2001) Wavellite, Fluellite and Minyulite from the Willard mine, Pershing County, Nevada: Mineralogical Record v. 32 no. 4 p. 297-304

Jensen M.C., Rota J.C. and Foord E.E. (1995) The Gold Quarry mine, Carlin-Trend, Eureka, Nevada: Mineralogical Record v. 26 no. 5 p. 449-474

Kolitsch U. (2004) Mg-rich Wolfeite (Fe,Mg)2(PO4)(OH) Structural refinement and Raman spectroscopic data: Act. Kristallogr., v. 59 no.9, p. 125-128

Mandrino J.A., Sturman B.D. and Corlett M.I (1977) Penikisite, the magnesium analogue of Kulanite from Yukon Territory; Canadian Mineralogist v. 15 p. 393-395

McCubbin E.M., Phillips B.L., Adcock C.T., Tait K.T., Steele A., Vaughn J.S. Fries M.D., Atudorei V., Vander Kaaden K.E. and Hausrath E.M. (2018) Discreditation of bobdownsite and the establishment of criteria for the identification of minerals with essential monofluorphosphate (PO3F+2): American Mineralogist v.103 no.8, p. 1319-1328

Megaw P.K.M. (2018) Santa Eulalia Mining District, Chihuahua, Mexico: Mineralogical Record v. 49 no.1 p. 4-184

Moore P.B. (1971) The Fe+23(H2O)n(PO4)2 Homologous Series: Crystal Chemical Relationships and Oxidation Equivalents: American Mineralogist, v. 56, Part 1 and 2, p. 1-17

Nriagu J. O. and Dell C.I (1974) Diagenetic Formation of Iron Phosphates in Recent Lake Sediments: American Mineralogist v. 59, p. 934-946

Roberts A.C., Cooper M.A., Hawthorne F.C., Gault R.A. Jensen M.C. and Foord E.E. (2003) Goldquarryite, a new Cd-bearing phosphate mineral from the Gold Quarry mine, Eureka County, Nevada: Mineralogical Record v. 34 no.3 p. 237-240

Robinson G.W., Van Velthuizen J, Ansell H.G. and Sturman B.D. (1992) Mineralogy of the Rapid Creek and Big Fish River Area Yukon Territory: Mineralogical Record v. 23 no.4 p. 4-47

Rothe M., Kleeberg A. and Hupfer M. (2016) The occurrence, identification and environmental relevance of vivianite in watherlogged soils and aquatic sediments: Earth Science Review v. 158, P. 51-64.

Sejkora J., Skovira J., Cejka J. and Plasil J. (2009) Cu-rich members of the beudantite-segnitite series from the Krupka Ore Disrtict, the Krusne hory Mountains, Czech Republic; Guornal of Geisciences v. 54, p. 355-371

Shannon E.V. (1923) On siderite and associated minerals from the Columbia River basalts at Spokane, Washington: Proceedings U.S. National Museum v. 62 art. 12.

Stoffregen R.E. and Alpers C. N. (1987) Woodhouseite and Svanbergite in hydrothermal ore deposits: products of apatite destruction during advanced argillic alteration: Canadian Mineralogist v. 25 p. 201-211

Sturman B.D. and Dunn P.J. (1984) Garyansellite, a new mineral from Yukon Territory, Canada: Canadian Mineralogist v. 64 p. 207-209

Tomes H. (2011) Arrojadite-(Ba,Na), a new Arrojadite-group mineral from the Big Fish River, Yukon Territory, Canada: M.Sc. Thesis, University of Toronto, 88 p.

Tomes, H., Tait, K., Nicklin, I., Peterson, R., and Beckett, R., (2013) Preliminary observations on the geology and mineralogy of the Rapid Creek Formation, Blow River and Davidson Mountains map area (NTS 117A/8 and NTS117A/9), Yukon. In: Yukon Exploration and Geology 2012, K.E. MacFarlane, M.G. Nordling, and P.J. Sack (eds.), Yukon Geological Survey, p. 147-161.

Wilson W.E. (2010) Famous Mineral Localities: The Clay Canyon Variscite mine, Fairfield, Utah: Mineralogical Record v. 41 no.4 p. 321-354

Wise W.S. (1990) The Mineralogy of the Mohawk Mine, San Benito County, California: San benito Museum Association Quarterly; v. 37(!), February 1990

Wise W. and Loh S.E. (1976) Equilibrium and origin of minerals in the system Al2O3-AlPO4-H2O: American Mineralogist v. 61 p. 409-413

Yeo G.M. (1992) Phosphorites, iron stones and secondary phosphates in mid-Cretaceous Flysch of the Blow Trough, Northern Yukon: In: Yukon Geology, V.3, Exploration and Geological Services Division, Yukon, Indian and Northern Affairs, Canada, p. 27-36

Young A. (2010) Famous Mineral Localities: The Balckbird mine, Lemhi County, Idaho: Mineralogical Record v. 41 no. 4 p. 355-372.

Appendix 1: New Chemical Analyses from the RRUFF database

The phosphate minerals of the Rapid Creek-Big Fish River area have been extensively analyzed in the RRUFF data base. The n= is the number of individual analyses to provide the sample average. The samples which have been analyzed in detail are:


Augelite: 2 average sample analyses posted with no significant difference between the sample chemistry n=8
1) Al2(P1.00O4)(OH0.98F0.02)3
2) Al2(P1.00O4)(OH)3

Brazilianite: there is no published chemical data for the Rapid Creek Brazilianite other than the SI saying it is almost pure.

Collinsite: 1 average sample with light and dark phases analyzed n=10
Ca2.00(Mg0.63Fe2+0.34Mn0.03)Σ=1(P1.00O4)2·2H2O ; = darker phase. The lighter phase is: Ca2.00(Mg0.58Fe2+0.38Mn0.04)Σ=1(P1.00O4)2·2H2O

Gormanite: 1 average sample posted with light and dark phases analyzed n=6
(Fe2+1.85Mg1.09Mn0.06)Σ=3(Al3.76Fe3+0.24)Σ=4(P1.00O4)4((OH5.65F0.35)Σ=6·2H2O ; (light phase) : (Fe2+1.53Mg1.42Mn0.05)Σ=3(Al3.89Fe3+0.11)Σ=4(P1.00O4)4((OH5.60F0.40)Σ=6·2H2O (dark phase)

Goyazite: 1 sample posted no analyses.

Kryzhanovskite-Garyansellite: 1 average sample analysis posted with no significant zonation n=10

Kulanite: 1 average sample analysis posted with no significant zonation n=10
(Ba0.97Sr0.03)Σ=1(Fe2+1.06Mn0.54Mg0.35Ca0.03Ti0.01 0.01)Σ=2(Al1.67Fe3+0.33)Σ=2(P1.00O4)3((OH)2.68F0.32)Σ=3; OH estimated by difference and charge balance

Lazulite 1 average sample analysis posted with no significant zonation n=12

Rapidcreekite: 1 sample posted no analysis

Vivianite 1 average sample analyses co-existing with maricite n=4
(Fe2+1.77Mg1.21Mn0.02)Σ=3(P1.00O4)2·8H2O; (dark phase, H2O is estimated by difference) : Na1.00Fe2+1.00P1.00O4 (light phase = maricite, with trace amounts of Mn)

Wardite 1 sample with only an average analysis n=1
(Na0.93 0.05Ca0.02)Σ=1(Al2.97Fe3+0.03)Σ=3(P1.00O4)2((OH)3.88F0.12)Σ=4·2H2O

Whitlockite: (still considered “Bobdownsite” even though it is discredited) 1 average sample analysis n=25
(Ca8.76Na0.24)Σ=9(Mg0.72Fe3+0.13Al0.11Fe2+0.04)Σ=1(P1.00O4)6(P1.00O3F1.00); Fe2+ and Fe3+ by charge balance


Arrojadite: 2 samples; 1(crystal) has no analysis and 2 (from the nodule locality) has an average sample analysis n=10
(Na1.80K0.20)Σ=2Fe2+1.00(Ca0.90Na2.10)Σ=3(Fe2+10.71Mg1.40Mn0.89)Σ=13(Al0.90Ti0.05Fe3+0.05)Σ=1(P1.00O4)11(P1.00O3OH)((OH)1.15F0.85)Σ=2; (OH) by stoichiometry and charge balance

Baricite: 1 average sample n=8
(Mg1.66Fe1.33Mn0.01)Σ=3(P1.00O4)2·8H2O; H2O estimated by difference

Ferroalluaudite: (nodule with maricite) 1 average sample analysis n=8

Maricite: 1 average sample analysis n=10

Satterlyite: 1 average sample analysis n=15
(Fe2+7.95Mg2.25Fe3+0.66Na0.66Mn0.48)Σ=12(P1.00O3OH)(P1.00O4)5((OH)5.00O1.00)Σ=6; OH estimated by charge balance

Wicksite: 1 sample no analysis

Wolfeite: 1 average sample analysis n=10
(Fe2+1.72Mg0.22Mn0.06)Σ=2P1.00O4((OH)0.87F0.13)Σ=1; (OH) estimated by difference and charge balance

As can be seen in both the crystals and nodule chemical analyses there are no pure end-member compositions in the Yukon Phosphates and several of the phases are intergrown with two different and separable phases in a single “crystal”. All of the analyzed phases mentioned have accompanying backscatter images most of which are homogenous on the grey scale but gormanite and collinsite have notable bi-shaded grey scales on the backscatter images with the light and dark grey shades analyzed. In all cases n is the total number of analyses of the phase.

Article has been viewed at least 1800 times.

Discuss this Article

6th Mar 2020 21:10 GMTDana Slaughter

Fabulous article---love the references and locality photos. The old proofreader in me is bothered by "nodual" rather than "nodule" but this is a small point (which I somehow could not let go unmentioned. My apologies!).

6th Mar 2020 21:37 GMTRichard Gunter Expert

Hi Dana:

Thanks for the compliment. Kim Tait from the ROM was kind enough to let me use her field photos.

The nodual vs nodule is an old debate. You find the former more common in the Canadian literature but they seem to be interchangeable.

6th Mar 2020 21:51 GMTDana Slaughter

Hi Richard,

Ha! I wondered if it might be something like that so I actually Googled the term but nothing came up. I really loved the article as it is chocked full of current information and wonderful locality photos. I would love to visit and would have to bring a good hammer,  a good camera and some bear treats. Thanks for sharing!

6th Mar 2020 22:18 GMTRichard Gunter Expert

Hi Dana:

A good rifle would be far more useful than bear treats. Those Barren Ground Grizzly Bears are aggressive and very dangerous. I spent a summer season working just south of Baker Lake, Nunavut and we had to be on the lookout for bears every time we did field reconnaissance.

6th Mar 2020 23:59 GMTKeith Compton Manager

Great article
A lot of really good info.

But bobdownsite photo is not quite “in Situ” but more in hand!!
Well done

7th Mar 2020 00:09 GMTAlfredo Petrov Manager

Thanks for the very useful article, Richard!

One addendum from last month‘s Tucson show, is that a couple of mislabelled (my fault) "augelites" from Rapid Creek turned out to be brazilianites when checked by Raman spectroscopy at the U of A. (A big thank you to Dave Joyce and Don Doell for bringing that to my attention.)
In the most recently dug lot of phosphates from there brazilianite seems to be at least as common as augelite, and their appearance is somewhat more similar to augelites than to the classic pegmatite brazilianites of Brazil, which makes things a bit confusing although, with practice, one can distinguish them by their crystal habit and striations.

7th Mar 2020 00:29 GMTRichard Gunter Expert

Hi Alfredo and Keith:

Thank you for the compliments.


I have found that several of the Brazilianite samples I have seen have frosted faces that Augelite never has. Other than that the two species look very similar. I am not surprised that the two are confused at Rapid Creek.

One thing I did notice was the Augelite crystals have end-to-end twinning that is manifest by changes in the orientation of the large faces. I have not seen that on Brazilianite.

7th Mar 2020 03:57 GMTAndrew Debnam

Great article Richard, 

7th Mar 2020 13:11 GMTEddy Vervloet Expert

Fantastic work, Richard!
Thank you for sharing all that valuable info!

7th Mar 2020 13:34 GMTJolyon Ralph Founder

I have reformatted the article slightly (changed headings and added in the table of contents in our 'official' manner), and fixed a couple of typos.

Thank you Richard!

7th Mar 2020 14:51 GMTTony L. Potucek Expert

Well done!  Thank you!  Photographs of the localities are always a plus.

8th Mar 2020 21:06 GMTSteve Rust Manager

Super article, really enjoyed reading it.

9th Mar 2020 14:32 GMTRichard Gunter Expert

Hi everyone.

Thank you for the compliments.

Jolyon, thank you for the changes.

9th Mar 2020 17:06 GMTFrank Ruehlicke

Wonderfull article Richard!  Important update for collectors of minerals from this site.

9th Mar 2020 19:54 GMTBranko Rieck Expert


thank you for this article. This is a blue-print of what an excellent article looks like.



9th Mar 2020 20:11 GMTLászló Horváth Manager

Thanks for the great update.

21st Mar 2020 14:47 GMTDavid K. Joyce Expert

Richard, Thanks so much for this! David KJ
Mineral and/or Locality  
Mindat.org is an outreach project of the Hudson Institute of Mineralogy, a 501(c)(3) not-for-profit organization. Public Relations by Blytheweigh.
Copyright © mindat.org and the Hudson Institute of Mineralogy 1993-2020, except where stated. Most political location boundaries are © OpenStreetMap contributors. Mindat.org relies on the contributions of thousands of members and supporters.
Privacy Policy - Terms & Conditions - Contact Us Current server date and time: April 9, 2020 02:46:14
Go to top of page