Mineralogical Magazine (2021), 85, 283–290 doi:10.1180/mgm.2021.40

Article

3+ Kernowite, Cu2Fe(AsO4)(OH)4⋅4H2O, the Fe -analogue of from , UK

Michael S. Rumsey1* , Mark D. Welch1, John Spratt2, Annette K. Kleppe3 and Martin Števko4,5 1Department of Earth Sciences, Natural History Museum, London SW7 5BD, UK; 2Core Research Laboratories, Natural History Museum, London SW7 5BD, UK; 3Diamond Lightsource UK, Harwell Science Park, Chilton, Oxfordshire OX11 0DE, UK; 4Earth Science Institute, Slovak Academy of Sciences, Dúbravská cesta 9, 840 05 Bratislava, Slovak Republic; and 5Department of Mineralogy and Petrology, National Museum, Cirkusová 1740, 193 00 Praha, Horní Počernice, Czech Republic

Abstract ⋅ The occurrence, chemical composition and structural characterisation of the new kernowite, ideally Cu2Fe(AsO4)(OH)4 4H2O, 3+ ⋅ the Fe -analogue of liroconite, Cu2Al(AsO4)(OH)4 4H2O, are described. Kernowite (IMA2020-053) occurs on specimens probably sourced from the mine, , Cornwall, UK, in the cavities of a quartz- rich in undifferentiated micro-crystalline grey sulfides and poorly crystalline phases including both pharmacosiderite and -group . The average compos- ition of kernowite determined from several holotype fragments by electron microprobe analysis is Cu1.88(Fe0.79Al0.09)Σ0.88(As1.12O4) ⋅ (OH)4 3.65H2O. The structure of kernowite has been determined in monoclinic space group I2/a (a non-standard setting of C2/c) by single-crystal X-ray diffraction (SCXRD) to R1 = 0.025, wR2 = 0.051 and Goodness-of-fit = 1.112. Unit-cell parameters from SCXRD are a = 12.9243(4) Å, b=7.5401(3) Å, c=10.0271(3) Å, β = 91.267(3)°, V=976.91(6) Å3 and Z = 4. The chemical formula of 3+ ⋅ this crystal indicated by SCXRD from refined site-scattering is Cu2(Fe0.84(1)Al0.16)AsO4(OH)4 4H2O. The network of hydrogen bonding has been determined and is similar to that reported for liroconite from Wheal Gorland by Plumhoff et al. (2020).

Keywords: kernowite, liroconite, Cornwall, iron aluminium arsenate, mineral collections, mineral museum, United Kingdom, new mineral (Received 14 April 2021; accepted 6 May 2021; Accepted Manuscript published online: 12 May 2021; Associate Editor: Juraj Majzlan)

History and occurrence analogue of liroconite, but the work was never formally written up. Kernowite was identified on an old museum specimen labelled as Due to the historic nature of the specimen, probably collected a liroconite from Wheal Gorland in the Sir Arthur Russell between 215–225 years ago, there is an element of uncertainty Collection of British Minerals housed at the Natural History with respect to its locational provenance, exacerbated in this Museum in London (Fig. 1). The specimen, registered as BM case as there has been informal debate for years questioning if 1964,R8908 was, prior to the ownership of the NHM and all old liroconite specimens are from the commonly attributed Russell before that, in the collection of the important Cornish Wheal Gorland. For lack of a means of verifying the exact source mineral collector and Member of Parliament, Philip Rashleigh location, the attributed locality of Wheal Gorland, given either by (1729–1811) and bears some notable resemblance to one of the Sir Arthur Russell or Philip Rashleigh is, henceforth, assumed to earliest figured examples of liroconite by James Sowerby in be correct. 1803 (Sowerby, 1804). Wheal Gorland was situated in the Parish of St. Day, Cornwall, The specimen was selected as being of interest during a sys- United Kingdom (50°14’30”N, 5°10’58”W) and was first recorded tematic study of liroconite by one of us (MSR) in the mid as a working mine in 1792 (Anon., 1799). It’s history since this 2000’s. Since its discovery, probably in the late 1790’s time is well documented, as is its fame within the mineral collect- (De Bournon, 1801) and approximate naming (as liriconite) by ing community for its rich, beautiful and largely unique suite of Jameson (1821) it has been regularly recognised as varying in col- arsenate minerals, including the iconic liroconite. The our from sky-blue to ‘verdigris’ or grass-green, speculation sug- locality and most of its mineral dumps no longer exist, having gested this was due to variation in phosphate–arsenate content been removed or levelled for recent housing developments. (Berry, 1938), but no modern confirmatory work had been per- Only a small dump, preserved as an SSSI (Site of Special formed. Specimen BM 1964,R8908 was selected for study due Scientific Interest) remains. to its unusual dark-green colour, more akin to that of an emerald Kernowite occurs with poorly-crystalline arsenic phases (Fig. 1). Chemical (energy-dispersive spectroscopy) analysis of a including pharmacosiderite and olivenite-group minerals in cav- small fragment suggested that the material was an Fe-dominant ities in a quartz-gossan, rich in undifferentiated micro-crystalline grey sulfides. Recent work by Plumhoff et al. (2020) suggests that the crystallisation of liroconite is related closely to the formation *Author for correspondence: Michael S. Rumsey, Email: [email protected] of arsenic-rich gels/mineraloids and it is suggested here, on the Cite this article: Rumsey M.S., Welch M.D., Spratt J., Kleppe A.K. and Števko M. (2021) ⋅ 3+ basis of the assemblage present, that kernowite is similar. It is Kernowite, Cu2Fe(AsO4)(OH)4 4H2O, the Fe -analogue of liroconite from Cornwall, UK. Mineralogical Magazine 85, 283–290. https://doi.org/10.1180/mgm.2021.40 interesting to note that the mineral and textural associations on

© The Author(s), 2021. Published by Cambridge University Press on behalf of The Mineralogical Society of Great Britain and Ireland 284 Michael S. Rumsey et al.

Table 1. Chemical data (n = 9) for kernowite.

Constituent Wt.% Range S.D. Standards

Fe2O3 15.38 14.53–15.97 0.51 Hematite CuO 36.00 35.36–37.22 0.63 Cu2O As2O5 31.12 28.25–33.30 0.26 NiAs (syn.) Al2O3 1.15 0.76–1.85 0.31 Corundum (syn.) CaO 0.02 bd–0.06 0.02 Wollastonite SiO2 0.09 0.04–0.14 0.03 Wollastonite H2Ocalc 24.69 24.52–24.92 0.82 Total 104.55 103.73–105.53 N/A

S.D. – Standard deviation; bd – below detection.

Gorland, providing context to the rarity of the kernowite–liroconite distinction. In a full study of over 60 samples, 25 of which were analysed and ten of which were ‘green’, only one with crystals exhibiting a 0.5 mm green rim and blue cores could be described as kernowite-containing and this was only in the green rims. This colour zonation from a blue core to green rim has been noted in liroconite specimens since at least 1803 (Sowerby) and is now deserving of more attention. Considering the many liroconite samples in existence, it is likely that other specimens bearing ker- nowite are present in public and private collections. However, we stress that accurate chemical analysis is essential for determin- ation, as very few of the ‘green-liroconites’ analysed thus far have been kernowite. Kernowite, is named after the word for Cornwall in the Cornish language, which is commonly spelt ‘Kernow’. As the exact locality attribution of Wheal Gorland, although likely, is Fig. 1. Holotype sample BM 1964,R8908, the large central crystal is ∼12 mm in size, not certain it is reasonable to name the mineral after the region all material used in the study was removed from the remnants of the cavity infill, that we know the sample is from. Furthermore, as the analogue probably the base of a broken crystal present on the left of the image. The large crys- of liroconite, kernowite is a doubly appropriate name, reflecting tals were not sampled so that their form on this historic specimen was preserved. liroconite’s place as one of Cornwall’s most iconic mineral species famous around the world for its striking colour, historic origin this sample are subtly different to the majority of specimens and rarity. Kernowite will join both and labelled liroconite from Wheal Gorland. as markers of the historic importance of Cornwall in mineral An extensive study by one of us (MS) recently identified a sciences. The region recognised by UNESCO as a world heritage second kernowite-containing specimen, also attributed to Wheal site based on its importance to the history of mining.

Fig. 2. Plot of the Al–Fe composition of 26 liroconite– kernowite specimens, the most Fe rich being the ker- nowite holotype. The plot additionally features an approximation of how the colour of the sample changes with increasing Fe content from blue through blue–green to green and dark green, many green and blue–green samples still fall in the compositional field of liroconite. Mineralogical Magazine 285

The mineral and name have been approved by the Table 2. Summary of information relating to data collection and structure International Mineralogical Association (IMA) Commission on refinement of kernowite. New Minerals, Nomenclature and Classification (IMA2020-053, Crystal data Rumsey et al., 2020). Holotype material is stored at the Natural Ideal chemical formula Cu2Fe(AsO4)(OH)4⋅4H2O ⋅ History Museum in London, within the Sir Arthur Russell Structural formula (SREF) Cu2Fe0.84Al0.16(AsO4)(OH)4 4H2O Collection of British Minerals, the sample consists of a specimen, Space group* I2/a a (Å) 12.9243(4) two polished microprobe blocks, a number of mounted crystal b (Å) 7.5401(3) fragments for single crystal studies and a small number of loose c (Å) 10.0271(3) tiny fragments, are all registered collectively under the number β (°) 91.267(3) BM1964,R8908. V (Å3) 976.91(6) Z 4 Reflections used for cell refinement 4171 Composition (I >7σI ) q range for cell measurement (°) 3.129−30.741 – All fragments subsampled for analysis from the holotype speci- Dx (Mg m 3) calc. 3.048 –1 men were taken from a cavity infill with no obvious crystal μ (mm ) calc. 8.983 form due to it completely encompassing the cavity. The largest Data collection ∼ Diffractometer Xcalibur E (1K Eos detector) fragment sub-sampled, 2 mm in size, was mounted in epoxy, Radiation, wavelength (Å) MoKα, 0.71073 polished and carbon coated. Fourteen chemical analyses were Crystal Green irregular transparent obtained on a Jeol JXA-8530A+ Hyper probe at the Natural tablet History Museum, London. Settings were 15 kV (beam voltage) Max. × med. × min. dimensions (mm) 0.166 × 0.117 × 0.080 μ Temperature (K) 293(1) and 10 nA (beam current) with a 10 m spot size. Cu, Al, Fe, Scan type, frame-width (°), frame-time (s) ω,1,30 Na, Si, Ca and As were sought, and other potentially relevant ele- Absorption correction Multi-scan (ABSPACK) ments, including P, Pb, Bi and Zn were ruled out through prior Tmin, Tmax 0.864, 1.227 analyses on a Cameca SX100. The data were corrected for any Reflection intensities collected 9520 Rσ 0.019 elemental overlaps and, like other studies on liroconite-series Independent reflections 1554 minerals (Plumhoff et al., 2020), showed dehydration under vac- Rint 0.026 uum, resulting in only the first nine points being used. It should Independent reflections with I >2σ(I ) 1419 θ θ be noted that elemental ratios remained consistent across all 14 min, max (°) 3.13, 31.46 analyses made. H O was calculated using data from the crystallo- Index range h ± 18, k ± 10, l ±14 2 Data completeness to 30°θ (%) 99.8 graphic study. Results are presented in Table 1, which correspond refinement to an average composition of Cu1.88(Fe0.79Al0.09)Σ0.88(As1.12O4) Independent reflections, restraints, 1555, 8, 96 ⋅ (OH)4 3.65H2O, the most Fe-rich analysis, normalising Al+Fe, parameters was Fe , the least Fe . Rint 0.026 93 83 σ The additional study utilised a Cameca SX100 electron R1[I >2 (I )], R1(all) 0.023, 0.027 wR2[I >2σ(I )], wR2(all) 0.050, 0.051 microprobe at the Laboratory of Electron Microscopy and GoF (F2) 1.112 Microanalysis of the Masaryk University and Czech Geological Restrained GoF (F2) 1.111 Survey in Brno, Czech Republic, operating in the wavelength- SHELX reflection weighting coefficients 0.0153, 4.620 dispersive (WDS) mode (15 kV, 20 nA and 10 μmwidebeam). a, b (Δ/σ)max <0.001 The following standards and X-ray lines were used to minimise – –3 Δρmax, Δρmin (e Å ) 0.62, –0.76 line overlaps: albite (NaKα), almandine (FeKα), fluorapatite (CaKα and PKα), Bi (BiMα), celestine (SKα), diopside *non-standard setting of C2/c (MgKα), gahnite (ZnKα), lammerite (CuLα and AsLα), sanidine α α α α α (AlK ,KK and SiK ), ScVO4 (VK ), spessartine (MnK ), topaz (FKα) and vanadinite (PbMα and ClKα). A total of 114 Neutral atomic scattering factors (Wilson, 1992) were used for all analyses from the 25 representative samples of liroconite atoms H, O, Al, Fe and Cu. (15blueand10withvariousshadesofgreen)aswellasdata Systematic absences were consistent with space groups C2/c for the holotype sample of kernowite are shown in an Fe vs. and Cc. Structure solution and refinement indicated that C2/c is Al plot (Fig. 2). the correct space group. To facilitate a comparison with the struc- ⋅ ture of the Al-analogue liroconite, Cu2Al(AsO4)(OH)4 4H2O, reported by Burns et al. (1991), we have used the non-standard Crystal structure setting I2/a and the data presented here were obtained by refine- A fragment of kernowite (0.165 mm × 0.115 mm × 0.080 mm) ment in this space group. We note that an attempt to determine removed from the cavity infill was attached to a non-diffracting the structure in non-centrosymmetric space group Ia (the other amorphous carbon fibre (0.01 mm diameter), itself glued to a possibility) resulted in high correlations between atoms related glass support rod (0.1 mm diameter), and turned out to be singu- by a centre of symmetry. Furthermore, the Flack parameter was larly crystalline. Data collection used an XcaliburE four-circle close to 0.5, indicative of true centrosymmetry and confirms diffractometer equipped with an Eos area detector and graphite- space group I2/a as the correct choice. monochromated MoKα radiation operated at 50 kV and 40 mA. Site scattering at the M3+ site was modelled using Al and Fe Raw reflection intensities were corrected for polarisation and neutral scattering factors and refined to 84(1)% Fe and 16% Al. Lorentz effects and converted to structure factors using the pro- As no other trivalent cations were detected by electron micro- gram CrysalisPro® (Rigaku Oxford Diffraction). Structure solution probe analysis, this ratio is considered to indicate the true occu- (Direct Methods) and refinement used SHELX (Sheldrick, 2015). pancy of this site. The site composition is also within the range 286 Michael S. Rumsey et al.

Table 3. Atom coordinates of kernowite (SG I2/a) and atom displacement parameters (Å2).

11 22 33 23 13 12 Site xyzUeq U U U U U U

As ¼ 0.04554(4) 0 0.01028(8) 0.0080(1) 0.0142(2) 0.0087(2) 0 0.0010(1) 0 Cu 0.13426(2) 0.28511(4) 0.77387(3) 0.01592(8) 0.0099(1) 0.0262(2) 0.0118(1) 0.0054(1) 0.0020(1) 0.0036(1) *M3+ 0 0 0 0.0127(2) 0.0081(2) 0.0202(3) 0.0098(3) –0.0015(2) 0.0006(2) –0.0016(2) O1 0.3545(1) –0.0848(2) 0.0085(2) 0.0164(3) 0.0108(8) 0.0169(9) 0.0217(9) 0.0027(7) 0.0020(6) 0.0023(6) O2 0.2556(1) 0.1753(3) 0.8624(2) 0.0157(3) 0.0092(7) 0.0265(9) 0.0117(8) 0.0067(7) 0.0021(6) 0.0039(7) OH1 0.0281(2) 0.4070(3) 0.6792(2) 0.0244(4) 0.0170(9) 0.043(1) 0.0140(9) 0.0088(8) 0.0055(7) 0.0147(8) OH2 0.0465(1) 0.2354(3) 0.9250(1) 0.0192(4) 0.0146(8) 0.026(1) 0.0167(8) 0.0057(7) 0.0047(6) 0.0043(7) OW1 0.1901(2) 0.6194(4) 0.8386(3) 0.0495(7) 0.040(1) 0.056(2) 0.053(2) –0.031(1) 0.005(1) 0.005(1) OW2 0.1083(2) 0.0067(4) 0.6231(2) 0.0319(5) 0.027(1) 0.038(1) 0.030(1) 0.002(1) 0.0089(9) –0.007(1) H1 –0.014(3) 0.438(6) 0.735(3) 0.050(9)** H2 0.005(3) 0.318(4) 0.909(4) 0.050(9)** H3 0.174(4) 0.701(5) 0.892(4) 0.08(1)** H4 0.188(4) 0.665(6) 0.763(2) 0.08(1)** H5 0.098(4) 0.078(6) 0.560(4) 0.08(1)** H6 0.171(2) –0.021(7) 0.620(5) 0.08(1)**

*Refined occupancy of M3+ site = 0.84(1) Fe + 0.16 Al. **Uiso

– of compositions determined by EMPA as Fe83 Fe93 (see above). Table 4. Bond lengths (Å) and polyhedral data for kernowite*. The occupancies of Cu and As sites were also allowed to vary: – – – their site-scattering values were consistent with full occupancy As O1 1.670(2) ×2 Cu OH1 1.890(2) Fe OH1 1.955(2) ×2 As–O2 1.694(2) ×2 Cu–OH2 1.950(2) Fe–O1 1.990(2) ×2 by Cu [0.95(1)] and As [0.96(1)], consistent with Al, Fe, Cu 1.682 Cu–O2 1.967(2) Fe–OH2 2.024(2) ×2 and As being the only significant cationic species detected by Volume 2.44 Cu–O2’ 2.017(2) 1.990 EMPA. Cu and As occupancies were set at 100% Cu and 100% BAV 0.761 Cu–OW2 2.604(3) Volume 10.50 – As for subsequent refinements. QE 1.000 Cu OW1 2.699(3) BAV 1.275 1.852 QE 1.001 All six non-equivalent H atoms were found after eight cycles of Volume 13.13 least-squares isotropic refinement. A soft restraint on the O–H BAV 66.51 bond-length of 0.85 ± 0.05 Å was used in the refinement of H QE 1.065 positions. A further soft restraint on the H⋅⋅⋅H distance (1.35 ± *Polyhedral volume (Å3), bond-angle variance (BAV, °2) and quadratic elongation (QE) 0.05 Å and consistent with the bond-length restraint) was used calculated according to Robinson et al. (1971) using VESTA (Momma and Izumi, 2011). to ensure a refined value of the H–O–H angle near to 104.5°, 3 the near-tetrahedral angle associated with sp bonding in H2O. Two H atoms are associated with a pair of non-equivalent OH Table 5. Bond-valence parameters (valence units) of kernowite calculated using Brown and Altermatt (1985). groups and four H with two non-equivalent H2O molecules. The Uiso values of H atoms of the hydroxyl and water molecules Cu *Al *Fe3+ As Σ oxygen were refined independently as a pair and a quartet, respectively. ×2↓ ×2↓ Full anisotropic refinement of all non-H atoms and isotropic O1 0.063 0.450 1.300 1.892 O2 0.860 ×2↓ 1.218 ×2↓ 2.078 ↓ refinement of all six H atoms gave final agreement indices R1 = OH1 0.565 0.069 0.495 ×2 1.129 ↓ 0.023, wR2 = 0.051 and Goodness-of-Fit = 1.112. Information OH2 0.481 0.057 0.410 ×2 0.948 relating to the data collection and structure refinement are sum- OW1 0.063 0.063 marised in Table 2. Atom coordinates and atom displacement OW2 0.082 0.082 Σ parameters are given in Table 3; bond-lengths and polyhedral cations 2.052 0.378 2.710 5.036 parameters in Table 4; bond-valences are shown in Table 5.A *Calculated for M3+ = 0.84 Fe3+ + 0.16 Al. list of structure factors and the crystallographic information file has been deposited with the Principal Editor of Mineralogical Magazine and is available as Supplementary material (see below). arranged in trans configuration and form two long Cu−O The structure of kernowite is shown in Fig. 3. It consists of bonds (2.604 Å and 2.699 Å). The hydrogen-bonding network 3+ – chains of alternating corner-sharing Fe O4(OH)2 octahedra of kernowite is shown schematically in Fig. 4 and donor acceptor ⋅⋅⋅ – – and AsO4 tetrahedra extending parallel to the a axis. A chain is OD(H) OA distances are given in Table 6. The H O H angles connected to two adjacent chains by CuO2(OH)2(H2O)2 dimers for both H2O molecules are 106(3)°. in which the pair of octahedra share an edge. CuO2(OH)2(H2O)2 dimers connect to an Fe–As chain by corner-sharing involving OH groups shared between Cu and Fe octahedra or an unproto- Powder X-ray diffraction nated oxygen between the Cu octahedron and As tetrahedron. The decision to use only a small quantity of kernowite for study precluded a standard flat-plate powder X-ray diffraction pattern being obtained. This was considered by the authors as an accept- Hydrogen bonding in kernowite able compromise as single-crystal data show that kernowite is Two of the six H atoms, H1 and H2, form OH groups whereas the clearly isostructural with its chemical analogue liroconite. A remaining H3–H6 atoms are associated with two non-equivalent quasi-randomised powder pattern using the crystal for which water molecules that form long Cu–O bonds. The latter are the structure was determined was obtained using a Mineralogical Magazine 287

Fig. 3. Crystal structure of kernowite. AsO4 tetrahedra, FeO2(OH)4 and CuO2(OH)2(H2O)2 octahedra are shown in purple, brown and blue, respectively. (a) The AsO4−FeO2(OH)4 chain extending parallel to [100]; (b) the fundamental structural polyhedral unit of kernowite that is polymerised to build the full structure; (c) a Cu2O2(OH)4(H2O)4 dimer showing the intra-dimer hydrogen bond between the two non-equivalent H2O molecules; (d) the structure of kernowite projected onto (001) and showing the chequerboard arrangement of polymerised structural units.

Gandolfi-like movement of w and ω circles on a Rigaku RAPID-II with those determined by single-crystal X-ray diffraction. A list diffractometer equipped with a cylindrical image-plate and of reflection data for kernowite obtained by the Gandolfi method graphite-monochromated CuKα radiation operated at 45 kV and described above is given in Table 7. 40 mA. The pattern was fitted using the Pawley method and refined unit-cell parameters (space group I2/a) obtained using Raman spectroscopy the program HighScore (®PanAnalytical Industries, Degen et al., 2014): a = 12.927(1), b=7.5430(9), c = 10.0332(7) Å, β = 91.212 The unpolarised Raman spectrum of kernowite was obtained by (4)° and V = 978.1(1) Å3. These values are in good agreement dispersing crystal fragments (<0.1 mm maximum size) on a 0.3 mm 288 Michael S. Rumsey et al.

instrument used was a Labram HR800 (Horiba Jobin–Yvon) spec- trometer on beamline I15 at Diamond Light Source, UK. It was equipped with 1200 g grating and a CCD detector. The spectra were excited by the 473 nm line of a 50 mW Cobalt Blues TM laser focused down to a 0.01 mm spot on the sample and collected through a 0.05 mm confocal aperture. The intrinsic resolution of − the spectrometer is <1 cm 1 and calibrations are accurate to ±1 − cm 1. The frequency of each Raman band was obtained by fitting Voigtian line profiles using a least-squares algorithm. The Raman – spectrum of kernowite is shown in Fig. 5 for the 100–1100 cm 1 – and 2950–3800 cm 1 ranges, where peaks occur. Peak assign- ments are based up on those for liroconite (Makreski et al., 2015). A Raman spectrum of liroconite from Wheal Gorland obtained from the RRUFF database (Lafuente et al., 2015), spec- trum ID R050535 is also shown in Fig. 5. There is a close corres- pondence between kernowite and liroconite Raman spectra. – – Modes between 300 cm 1 and 800 cm 1 can be associated with vibrations of the AsO4 tetrahedron and Fe/Al and Cu – Fig. 4. Schematic diagram of the hydrogen-bonding network in kernowite. Oxygen octahedra. The strong doublet 854, 871 cm 1 in the kernowite donor−acceptor distances (Å) are shown with an arrow pointing from the donor to the acceptor. ‘COS’ indicates a centre of symmetry that acts upon the network spectrum (not resolved in the RRUFF liroconite spectrum) corre- topology. sponds to the symmetric and antisymmetric stretching modes of the AsO4 tetrahedron. – – –1 Table 6. Oxygen donor−acceptor distances O (H)⋅⋅⋅⋅O in kernowite (this study) The O H stretching region (2950 3800 cm )comprisesa D A –1 and those of liroconite (Plumhoff et al., 2020). The correspondence for atom broad feature centred on 3280 cm ,andasharppeakwitha labels of the two studies is kernowite/liroconite: O1/O1, O2/O2, OH1/O3, weaker shoulder at a higher wavenumber. The OH spectrum OH2/O4, OW1/O5 and OW2/O6. is fitted best with a four-peak model in which the broad feature – – comprises two similar peaks at 3224 cm 1 and 3363 cm 1,and Kernowite (this study) Liroconite (Plumhoff et al., 2020) – – two narrow peaks at 3470 cm 1 and 3557 cm 1. Following ⋅⋅⋅ ⋅⋅⋅ Donor Acceptor d(OD OA) Å Donor Acceptor d(OD OA)Å Makreski et al. (2015) we assign the two broad peaks to OH1 OW2 2.785 O3 O6 2.789 the two non-equivalent H2O groups, and the two narrow OH2 OW2 2.895 O4 O6 2.909 peaks to the two non-equivalent hydroxyl groups. The two OW1 O1 2.774 O5 O1 2.795 – narrow peaks are assigned to O H vibrations of non-H2O OW1 OW1* 3.091 O5 O1* 3.472 hydroxyl groups. In addition to the broad modes centred on OW2 OW1 2.792 O6 O5 2.775 –1 3280 cm (due to H2O), the RRUFF Raman spectrum of liro- OW2 OH2 2.881 O6 O4 2.954 – conite has a single narrow peak at 3518 cm 1,avaluethatlies * Different acceptor atoms are reported for the two minerals. See text for discussion. between the wavenumber values of the two narrow peaks of kernowite. Table 7. Quasi-powder diffraction data (d in Å, I in %) for kernowite collected Provisionally, we interpret the two narrow peaks of the kerno- from the single crystal used in the structure determination. The calculated wite Raman spectrum as being due to OH for the bridging con- intensities are based upon the structure determination.* figurations Cu–O(H)–Fe and Cu–O(H)–Al. The much lower –1 –1 I/Imax obs I/Imax calc dhkl obs dhkl calc hkl intensity of the 3557 cm peak compared with the 3470 cm peak, is then consistent with the minor Al content of kernowite. 100 100 6.560 6.515 1 1 0 Composition-sensitive vibrational frequencies for O–H groups 91 84 6.067 6.029 0 1 1 28 25 3.970 3.954 1 1 2 are well-documented for many minerals, most notably amphi- 10 10 3.755 3.741 3 1 0 boles and micas. An alternative interpretation is that the two nar- 18 13 3.411 3.399 1 2 1 row OH modes are due to the two crystallographically 41 28 3.066 3.057 0 1 3 non-equivalent O–H groups. However, this interpretation (1) 33 46 3.035 3.014 0 2 2  assumes that there is no compositional effect; and (2) ignores 19 11 2.871 2.863 4 1 1 ⋅⋅⋅ ⋅⋅⋅ 30 22 2.841 2.833 4 1 1 the fact that O(H1) O and O(H2) O distances are similar 16 22 2.750 2.743 2 1 3 (2.8 Å and 2.9 Å), i.e. the hydrogen bonds are likely to be of simi- 27 22 (13, 9) 2.728 2.719, 2.690 2 2 2, 4 0 2 lar strength, implying similar bond-lengths/bond-strengths for 20 14 2.460 2.454 4 2 0 the two O–H groups. However, we recognise that this point 15 13 2.216 2.211 1 3 2 13 14 2.185 2.181 5 1 2 needs further study. Nonetheless, the wavenumber difference 13 12 (7, 5) 2.105 2.101, 2.093 3 1 4, 5 2 1 between the narrow peaks of kernowite and liroconite is clear 19 14 1.771 1.768 4 2 4 and may be of diagnostic value in recognising Fe-rich and 11 12 (7, 5) 1.719 1.740, 1.733 4 2 4, 5 1 4 Al-rich crystals.

*Only observed reflections with I/Imax ≥ 10% are listed (2θmax = 52.8°). The five strongest reflections are shown in bold. Discussion diameter culet of an ultra-low fluorescence Raman diamond Plumhoff et al. (2020) determined the structure of liroconite of ⋅ (Almax Industries). Ambient spectra were collected in 180° back- composition Cu2Al(AsO4)0.86(PO4)0.14(OH)4 4H2O from Wheal − scattering geometry over the range 100–4000 cm 1. The Gorland in space group I2/c with the same unit-cell setting as Mineralogical Magazine 289

Fig. 5. Ambient unpolarised Raman spectrum of a single crystal of kernowite. The fits of the two- and four-peak models for the OH-stretching region are shown. The Raman spectrum of liroconite from the RRUFF database is also shown. See text for discussion. 290 Michael S. Rumsey et al.

I2/a [a = 12.643(1), b=7.5684(7), c = 9.880(1) Å, β = 91.276(8)° References and V = 945.1(2) Å3]. This liroconite is nearly identical topologic- Anon. (1799) Report from the Committee appointed to enquire into the state of ally to liroconite with space group I2/a (Burns et al., 1991)and the copper mines and copper trade of this Kingdom. Ordered to be printed kernowite. Its hydrogen-bonding configuration (Table 6)is 7th May 1799. Parliamentary Papers printed by order of the House of essentially analogous to that of kernowite, except for one H Commons, 1731 to 1800, vol. 52. atom, their H2O(5), for which Plumhoff et al.,foundavery Berry L.G. (1938) Observations on conichalcite, cornwallite, euchroite, liroco- ⋅⋅⋅⋅ 36 – long associated OD OA distance of 3.47Å. This H atom nite and olivenite. American Mineralogist, , 484 503. does not appear to be directed towards any plausible hydrogen- Brown I.D. and Altermatt D. (1985) Bond-valence parameters obtained from a bond acceptor. The corresponding atom in kernowite (H4) systematic analysis of the inorganic crystal structure database. Acta B41 – shows a clear orientation towards oxygen OW1, thereby provid- Crystallographica, , 244 247. Burns P. C., Eby R.K. and Hawthorne F.C. (1991) Refinement of the struc- ing connectivity between the two OW1 H2O molecules. The ⋅⋅⋅ ture of liroconite, a heteropolyhedral framework oxysalt mineral. Acta OW1 OW1 distance in kernowite is 3.09 Å, compared with C47 – ⋅⋅⋅⋅ Crystallographica, , 916 919. the corresponding O5 O5 distance of 3.18 Å in liroconite de Bournon J-L. Count (1801) Description of the arseniates of copper and of reported by Plumhoff et al. (2020), which is almost the same iron from the County of Cornwall. Philosophical Transactions, 91, 169–192. as 3.17 Å of Burns et al. (1991), who did not locate H atoms Degen T., Sadki M., Bron E., König U. and Néner G. (2014) The HighScore for their liroconite crystal. The shorter OW1⋅⋅⋅OW1 (O5⋅⋅⋅O5) suite. Powder Diffraction, 29 (Supplement 2), S13–S18. http://doi.org/ distance in kernowite may be responsible for the formation of 10.1017/S0885715614000840 a more directional hydrogen bond for the H4 atom in this min- Jameson R. (1821) Manual of Mineralogy. A. Constable, Edinburgh pp.491. eral compared with H O(5) in liroconite. This subtle difference Lafuente B., Downs R.T., Yang H. and Stone N. (2015) The power of databases: 2 – between the structures of kernowite and liroconite requires the RRUFF project. Pp 1 30 in: Highlights in Mineralogical Crystallography further elucidation. (T Armbruster and RM Danisi, editors). W. De Gruyter, Berlin, Germany. Makreski P., Jovanovski S., Pejov L., Petruševki G., Ugarković S. and An exploration of the extent of Al–Fe3+ solid solution and its Jovanovski G. (2015) Theoretical and experimental study of the vibrational structural systematics in liroconite–kernowite would seem to be ⋅ spectra of liroconite, Cu2Al(AsO4)(OH)4 4H2O and bayldonite, Cu3Pb[O worthwhile. Thus far, it appears that the compositions of natural 79 – (AsO3OH)2(OH)2]. Vibrational Spectroscopy, ,36 43. liroconite and kernowite are binary with respect to Al and Fe and Momma K. and Izumi F. (2011) VESTA 3 for three-dimensional visualization of the intensity of green colouration is a reasonable indicator for crystal, volumetric and morphology data. Journal of Applied Crystallography, samples worthy of further analysis. The substitution of P for 44,1272–1276. As at the tetrahedral site is a further aspect of the structure Plumhoff A.M., Dachs E., Benisek A., Plášil J., Sejkora J., Števko M., Rumsey M.S. that would benefit further study. In view of the success of and Majzlan J. (2020) Thermodynamic properties, crystal structure and phase ⋅ Plumhoff et al. (2020) in synthesising liroconite crystals large relations of pushcharovskite Cu(AsO3OH)(H2O) 0.5H2O, geminite Cu ⋅ enough for structure determination by single-crystal XRD, syn- (AsO3OH)(H2O) and liroconite Cu2Al(AsO4)(OH)4 4H2O. European Journal of Mineralogy, 32, 285–304. theses of a series of Al–Fe and As–P phases would seem to be Robinson K.,GibbsG.V. and Ribbe P.H. (1971)Quadraticelongation: a quantitative worth undertaking in order to quantify the crystal-chemistry measure of distortion in coordination polyhedra. Science, 172, 567–570. of the solid solution. Rumsey M.S., Welch M.D., Spratt J., Kleppe A. and Števko M. (2020) Kernowite, IMA 2020-053. CNMNC Newsletter No. 58; Mineralogical Acknowledgments. We thank Roy Starkey for clarifying information per- Magazine, 84, 971–975, https://doi.org/10.1180/mgm.2020.93 taining to the Sir Arthur Russell collection of British Minerals, Ed Loye for Sheldrick G.M. (2015) Crystal structure refinement with SHELX. Acta discussion and Callum Hatch and Tony Wighton of the NHM Preparation Crystallographica, C71,3–8. laboratories for creating the holotype probe blocks. Sowerby J. (1804) British Mineralogy, Volume I. R.Taylor & Co, London, pp. 224. Wilson A.J.C. (editor) (1992) International Tables for Crystallography, Volume C: Supplementary material. To view supplementary material for this article, Mathematical, Physical and Chemical Tables. Kluwer Academic, Dordrecht, please visit https://doi.org/10.1180/mgm.2021.40 Netherlands.