LETTER

Vanadium-rich ruby and sapphire within Mogok Gemfield,Myanmar: implications for gem color and genesis

Khin Zaw & Lin Sutherland & Tzen-Fu Yui &Sebastien Meffre & Kyaw Thu

Received: 29 December 2012 /Accepted: 25 July 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract Rubies and sapphires are of both scientific andcommercial interest. These gemstones are corundum coloredby transition elements within the alumina crystal lattice: Cr3+

yields red in ruby and Fe2+, Fe3+, and Ti4+ ionic interactionscolor sapphires. A minor ion, V3+ induces slate to purplecolors and color change in some sapphires, but its role incoloring rubies remains enigmatic. Trace element and oxygenisotope composition provide genetic signatures for naturalcorundum and assist geographic typing. Here, we show thatV can dominate chromophore contents in Mogok ruby suites.This raises implications for their color quality, enhancementtreatments, geographic origin, exploration and exploitationand their comparison with rubies elsewhere. Precise LA-ICP-MS analysis of ruby and sapphire from Mogok placerand in situ deposits reveal that V can exceed 5,000 ppm,giving V/Cr, V/Fe and V/Ti ratios up to 26, 78, and 97respectively. Such values significantly exceed those foundelsewhere suggesting a localized geological control on V-

rich ruby distribution. Our results demonstrate that detailedgeochemical studies of ruby suites reveal that V is a potentialruby tracer, encourage comparisons of V/Cr-variation be-tween ruby suites and widen the scope for geographic typingand genesis of ruby. This will allow more precise comparisonof Asian and other ruby fields and assist confirmation ofMogok sources for rubies in historical and contemporarygems and jewelry.

Keywords Ruby . Sapphire . Color . LA-ICP-MS .

Vanadium .Mogok .Myanmar

Introduction

Rubies (red corundum) are mined in several countries, includ-ingMyanmar, Thailand, Cambodia, Vietnam, Kenya, Afghan-istan,Madagascar, Pakistan, Sri Lanka, and Tanzania (Hughes1997; Graham et al. 2008; Garnier et al. 2008). The “pigeon-blood” rubies from Mogok, Myanmar, have been regarded asthe finest in the world for the last centuries. Corundum(Al2O3) becomes red when it contains sufficient trace amountsof the chromophore chromium (III) oxide, which gives rubiestheir highly saturated, deep red color. When iron and titaniumare the principal chromophores in corundum, the resultingblue gem is a sapphire. For rubies, as with many gemstones,color (as defined by hue, saturation, and tone) is a vital factorin determining value. The hue is divided into primary, sec-ondary, and possibly tertiary hues, with red the most impor-tant. Among secondary hues such as orange, pink, purple, andeven violet, purple is preferred. This hue deepens the primaryred hue, as in top-quality Mogok rubies. Here, we consider therole of V in affecting colors, especially the hues of Mogokrubies, which are the most prized in the world. We show fromprecise laser ablation inductively coupled plasma mass spec-trometry (LA-ICP-MS) analyses and imaging of Mogok ruby/

Editorial handling: C. Li and G. Beaudoin

K. Zaw (*) : S. MeffreCODES ARC Centre of Excellence in Ore Deposits, University ofTasmania, Hobart, TAS, Australiae-mail: khin.zaw@utas.edu.au

L. SutherlandSchool of Science & Health, University of Western Sydney, Penrith,NSW 2751, Australia

L. SutherlandGeoscience, Australian Museum, 6 College Street, Sydney, NSW,Australia

T.<F. YuiInstitute of Earth Sciences, Academia Sinica, Nankang, Taipei,Taiwan

K. ThuGeology Department, Yangon University, Yangon, Myanmar

Miner DepositaDOI 10.1007/s00126-014-0545-0

sapphire samples that V can take a dominant role in chromo-phore content in Mogok rubies. Here, we exploregemmological, geographic typing, gem exploration, and gemexploitation aspects of V in ruby suites.

Geological setting

Myanmar can be sub-divided into six N–S trending majortectonic domains from west to east: (1) the Arakan(Rakhine) Coastal Strip is an ensimatic fore-deep, (2) theIndo-Burman Ranges represent an outer-arc or fore-arc, (3)the Western Inner-Burman Tertiary Basin is considered to bean inter-arc basin, (4) the Central Volcanic Belt (CentralVolcanic Line) represents an inner magmatic-volcanic arc,(5) the Eastern Inner-Burman Tertiary Basin formed as aback-arc basin and (6) the Shan-Tenasserim Massif occurs asensialic, Sino-Burman Ranges. The Sagaing Fault forms atectonically significant boundary between the Eastern Inner-Burman Tertiary Basin (back-arc basin) and the continental,ensialic Sino-Burman Ranges (Bender 1983; Khin Zaw 1989,1990) (Fig. 1).

The Arakan coastal strip comprises Miocene molassic sed-imentary rocks that extend northward into the Assam Basin ofthe Northeastern Indian Ocean. The Indo-Burman Rangesform an outer-arc or fore-arc that extends south from theHimalayan Frontal Range (~96° E, 27° N) to the Bay ofBengal (~94.3° E, 27° N). These rocks are underlain byTriassic turbidites and dismembered ophiolites in the east,and tightly folded turbidites of Cretaceous to Early Eoceneage in the west (Clegg 1941; Searle and Haq 1964). The rocksin this belt typically dip towards the east and strike parallel tothe N–S trend of the ranges. The central lowlands and centralBasin contain a thickness of up to 15 km of dominantlyTertiary marine and fluviatile sedimentary rocks and are di-vided into the Western Inner-Burman Tertiary Basin (inter-arctrough) and the Eastern Inner-Burman Tertiary Basin (back-arc trough). Both basins are separated by the Inner Magmatic-Volcanic Arc that contains the Cretaceous granitoid plutonsandMiocene volcano-sedimentary rocks hosting the high- andlow-sulfidation epithermal Cu ± Au deposits at Monywa incentral Myanmar (Khin Zaw et al. 2014). The Shan-Tenasserim Massif is also known as the Eastern Highlandsor the Sino-Burman Ranges (Bender 1983). These ranges areunderlain by Paleozoic sedimentary rocks with minor Meso-zoic clastic rocks. The oldest rocks in this area are thought tobe Late Precambrian sedimentary rocks, although their agehas not yet been confirmed by radiometric dating. They con-sist of weakly metamorphosed, turbiditic, clastic rocks of theChaung Magyi Group. The Paleozoic rocks are both clasticand calcareous sedimentary rocks with minor volcanic rocks.The granitic belt occurs in the eastern and western areas of theSino-Burman Ranges, immediately east of the Sagaing Fault.The eastern granitoids are probably Triassic or older as they lie

along strike from the Triassic I-S-type granites of northernThailand (Khin Zaw 1990) and the granitic belt of the westernMain Range of Peninsula Malaysia.

Recent U-Pb SHRIMP dating of igneous zircons yieldedages of Early Cretaceous to Eocene for the western granites,whereas the age of central granite plutons ranged from Juras-sic through Cretaceous to Miocene (Barley et al. 2003). Thecentral granites belong to the Mogok Metamorphic Belt(MMB) which is a 1,000 km long arcuate belt along thewestern margin of the Shan-Tenasserim Massif, east of theSagaing Fault (Barley et al. 2003; Searle et al. 2007; Mitchellet al. 2004, 2007). The belt extends from southern Myanmarthrough Yebokson, Kyaukse, Mandalay, to Mogok and thenbends westward to the eastern Himalayas (Fig. 1).

TheMogok area (Iyer 1953; Hughes 1997; Themelis 2008)hosts world-class deposits of rubies, sapphires and other gem-stones and lies at the northeastern end of the MMB. Radio-metric dating of magmatic and metamorphic rocks along theMogok Belt includes an Ar–Ar biotite age of 15.8 Ma for theKabaing granitoid and Ar–Ar biotite ages of three nearbymetamorphic rocks ranging from 16.5 to 19.5 Ma (Bertrandet al. 2001). These ages, along with Ar–Ar phlogopite ages fora ruby-bearing marble of 18.7 Ma and for two ruby-freemarbles of 17.1 and 17.9 Ma (Garnier et al. 2006), all appearto be resetting older ages due to later tectonic and metamor-phic affects, as U–Pb dating of zircon inclusions in a Mogokruby gave an age of 31 to 32Ma (Khin Zaw et al. 2008, 2010).Kyaw Thu (2007) dated leucocratic granite intrusions atMogok using the LA-ICP-MS U-Pb zircon method and ob-tained ages of 45 and 25 Ma. These ages suggest that theMMB has experienced multiple tectonic magmatic and meta-morphic, uplift and exhumation processes, most likely accom-panied by metamorphic core complex formation from as earlyas the Jurassic to recent times.

The Mogok area consists of a series of undifferentiatedhigh-grade metamorphic rocks. The dominant unit is bandedgneiss, with biotite, garnet, sillimanite and oligoclase. It isinterspersed with quartzite and bands and lenses of marble(Fig. 2) (Khin Zaw 1990, 1998; Kyaw Thu 2007; Themelis2008). Marbles mainly contain calcite, while dolomitic mar-bles are limited. Some marbles contain ruby, whereas othershave spinel, forsterite or diopside, or phlogopite-graphite mar-bles (Kammerling et al. 1994; Kyaw Thu 2007; Themelis2008). Rutile, graphite, calcite, pargasite, zircon, titanite, ap-atite, scapolite, spinel and zircon also form inclusions inMogok rubies (Kammerling et al. 1994; Thu 2007; Themelis2008; Khin Zaw et al. 2008, 2010). The metamorphic rocksare associated with alkaline rocks (mostly sodic nepheline–syenite and syenite–pegmatite) and leucogranites, accompa-nied by a mafic–ultramafic suite of peridotite, minor gabbro,and norite. Biotite granites are widely exposed at Kabaing andThabeikkyin, near Mogok and some syenite pegmatites con-tain sapphires (Khin Zaw 1990, 1998; Waltham 1999;

Miner Deposita

Fig. 1 Regional setting and locations map of Mogok area, Myanmar(modified after Bender 1983; Khin Zaw 1989, 1990). (1) The Arakan(Rakhine) Coastal Strip, (2) the Indo-Burman Ranges, (3) the Western

Inner-Burman Tertiary Basin, (4) the Central Volcanic Belt (CentralVolcanic Line), (5) the Eastern Inner-Burman Tertiary Basin, and (6)the Shan-Tenasserim Massif

Miner Deposita

Themelis 2008; Kyaw Thu 2007). Late-stage pegmatites andaplites commonly intrude the Kabaing granitoids andmetasedimentary rocks (Khin Zaw 1998).

Methodology

In this study, LA-ICP-MS analysis and imaging were used todetermine the trace element composition of rubies and sap-phires from Mogok gemfield, as shown in Table 1. Theanalyses were performed at CODES, University of Tasmania,using a New Wave 193 nm solid state laser coupled with anAgilent 7500 cs quadrupole inductively coupled plasma massspectrometer (ICP-MS). Ablation was performed in a Heatmosphere in a custom-made small volume ablation cellusing NIST610 as the primary standard and BCR2G as thesecondary standard. The laser was pulsed at 10 Hz on a

100 μm spot size with a fluence of 2 J cm−2. Each analysisconsisted of 30 s of background gas, followed by 30 s ofablation time counting for 10 ms on each mass. Standard datareduction procedures were used for the spot data using alumi-num as the internal standard element (e.g., see Large et al.2009). The images were processed using similar methodsexcept that normalization was based on the average Al forthe entire image. Based on the normalization, a count/ppmvalue was then applied to each pixel for each element toprovide a quantified map of the corundum. O-isotope analyseswere performed by the CO2 laser-fluorination method (Yui2000), at the Institute of Earth Sciences, Academia Sinica,Taipei. A Finnigan MAT 252 mass spectrometer wasemployed to analyze the CO2 gas. The results are reportedas per mil 18O values relative to Vienna StandardMean OceanWater (V-SMOW). The analyzed sample weight ranges from1.8 to 2.9 mg and the analytical precision is slightly better than0.1‰ based on 17 analyses of the UWG-2 garnet standard

Fig. 2 Map showing sample location, general distribution of the ruby/sapphire mine/workings and geological setting of the Mogok gemfield,Myanmar (modified using Google map and after Kyaw Thu 2007;Themelis 2008). The location of studied samples of the gem depositsand other localities of ruby/sapphire mine/workings were plotted usingGoogle map and hand-held GPS. (1) Sin-khwa, (2) Lone-sho, (3) Pingu-

taung, (4) Htin-shu-taung, (5) Ohnbin-ywe-htwet, (6) Baw-mar, (7)Kadoke-tat, (8) Mye-me, (9) Chaung-gyi, (10) Kolan, (11) Baw-lone-lay, (12) Baw-padan, (13) Win-hta-yan, (14) Pan-taw, (15) Kyauk-war,(16) Ye-bauk-tayar, (17) Le-shu-kone-zan (not shown on themap; locatedabout 7 km east of Ohnbin-ywe-htwet), (18) Pazaun-seik, (19) Lay-thar,(20) Lay-thar

Miner Deposita

Table 1 Location (latitude and longitude) and summary of LA-ICP-MS trace element contents of ruby and sapphire samples from theMogok gemfield,Myanmar. The sample locations are shown in Fig. 2

Sample site Color Latitude Longitude No. V Cr Fe Ti Mg Ga V/Cr V+Cr/Ga

High V (>90 pm) metamorphic suites

Sin-khwa (1)a Red 22° 54′ 31″ 96° 23′ 10″ 1.r 5,044 369 76 56 39 27 13.67 200.48

1.c 5,213 433 67 54 34 31 12.04 182.13

Sin-khwa (1)a Red 22° 54′ 31″ 96° 23′ 10″ 2.r 3,960 3,758 79 132 105 90.1 1.05 85.66

2.c 4,271 165 59 95 31 27.6 25.89 160.72

Lone-sho (2)a Red 22° 54′ 34″ 96° 23′ 30″ 1.r 2,970 867 65 76 60 52.4 3.43 73.23

1.c 101 9 54 26 36 27.2 11.22 4.04

Pingu-taung (3)b Red-purple 22° 54′ 19″ 96° 24′ 05″ 1.r 906 1,039 59 95 30 75.8 0.87 25.66

1.c 1,297 1,229 57 34 27 69 1.06 36.61

Pingu-taung (3)b Purple 22° 54′ 19″ 96° 24′ 05″ 2.r 456 2,896 67 134 72 92.3 0.16 36.32

2.c 575 1,154 58 1,580 68 64.7 0.5 26.72

Htin-shu-taung (4)a Purple 23° 00′ 23.5″ 96° 28′ 40″ 1.r 153 521 1,228 50 15 107.8 0.29 6.25

1.c 148 549 1,232 46 11 109.9 0.27 6.34

Htin-shu-taung (4)a Dark purple 23° 00′ 23.5″ 96° 28′ 40″ 2.r 465 1,951 901 150 57 89.4 0.24 26.92

2.c 341 1,691 955 105 66 84.6 0.2 24.02

Ohnbin-ywe-htwet (5)a Light pink 22° 56′ 32″ 96° 32′ 10″ 1.r 119 1,071 28 106 61 112.7 0.11 10.56

1.c 104 1,005 27 117 81 106.3 0.1 10.43

Ohnbin-ywe-htwet (5)a Light pink 22° 56′ 32″ 96° 32′ 10″ 2.r 181 524 51 140 88 102.4 0.35 6.88

2.c 174 714 49 92 78 87.5 0.24 10.15

Baw-mar (6)a Pink 22° 55′ 39.9″ 96° 24′ 43″ 1.r 96 908 671 248 144 107.4 0.11 9.35

1.c 98 928 655 246 135 109.5 0.11 9.37

Baw-mar (6)a Red 22° 55′ 39.9″ 96° 24′ 43″ 1.r 186 1,560 131 335 215 64.5 0.12 27.07

1.c 206 1,250 138 348 260 66.9 0.17 21.76

Kadoke-tat (7)b Pink 22° 56′ 02″ 96° 28′ 01″ 1.r 171 165 88 47 25 94.7 1.04 3.55

1.c 163 277 83 57 35 104 0.59 4.23

Mye-me (8)a Red 22° 56′ 02″ 96° 26′ 42″ 1.r 520 1,221 77 95 54 78.7 0.43 22.12

1.c 483 356 60 88 50 80.7 1.36 10.4

Mye-me (8)a Purple-pink 22° 56′ 02″ 96° 26′ 42″ 2.r 1,555 185 102 100 53 41 8.41 42.44

2.c 1,430 154 120 59 32 12.3 9.29 128.8

Chaung-gyi (9)a Red 22° 57′ 51″ 96° 30′ 32″ 1.r 92 353 110 85 48 7.5 0.26 59.33

1.c 99 313 <52 224 57 7.7 0.32 3.51

Chaung-gyi (9)a Red 22° 57′ 51″ 96° 30′ 32″ 2.r 117 28 83 157 23 179 4.18 0.81

2.c 228 712 123 31 15 185 0.32 5.08

Kolan (10)b Red 22° 55′ 55″ 96° 26′ 33″ 2.r 226 802 396 143 69 148 0.28 6.94

2.c 223 683 388 125 61 155 0.32 7.78

Baw-lone-lay (11)b Red 22° 54′ 38″ 96° 24′ 07″ 1.r 257 1,671 129 87 45 164 0.15 11.76

1.c 191 1,871 148 109 52 119 0.1 17.33

Baw-lone-lay (11)b Red 22° 54′ 38″ 96° 24′ 07″ 2.r 201 1,810 100 103 53 69.9 0.11 28.77

2.c 124 1,867 67 152 69 54.8 0.07 36.33

Baw-padan (12)b Red 22° 56′ 09″ 96° 27′ 05″ 1.r 206 29 442 70 33 106 7.1 2.22

1.c 347 2,339 270 74 40 166 0.15 16.18

Baw-padan (12)b Red 22° 56′ 09″ 96° 27′ 05″ 2.r 109 3,359 310 69 35 142 0.03 24.42

2.c 96 3,003 <45 84 40 131 0.03 22.92

Win-hta-yan (13)b Red-purple 22° 55′ 34″ 96° 26′ 12″ 1.r 202 1,327 <61 292 156 89.8 0.15 17.03

1.c 193 1,254 <47 279 157 88.7 0.15 16.31

Win-hta-yan (13)b Red-purple 22° 55′ 34″ 96° 26′ 12″ 2.r 340 2,081 173 93 52 80.2 0.16 30.19

2.c 261 297 <45 7,639 63 60.4 0.88 9.24

Average 13 sites 46 759 1,103 224 153 66 88.5 2.35 33.21

Low V (<35 ppm) intermediate suites

Lone-sho (2)a Blue 22° 54′ 34″ 96° 23′ 30″ 2.r 3 <1.1 679 90 59 77.8 >2.54 0.05

2.c 5 <1.0 849 209 133 75.5 >4.48 0.08

Pan-taw (14)a Light blue 22° 57′ 36.7″ 96° 24′ 58.9″ 1.r 7 1 609 138 71 93.6 5.34 0.09

Miner Deposita

(Yui 2000), which was employed to normalize the daily dataand has a suggested δ18O value of +5.8‰ (Valley et al. 1995).Detailed oxygen isotope data are shown in Table 2.

Trace element composition

LA-ICP-MS analysis and imaging indicate that significantvariations exist in V (92–5,045 ppm) and Cr (28–3,760 ppm) in the rubies and pink to purple sapphires and thatthey generally exceed Fe (27–1,232 ppm) and Ti values(mostly <350, but 1,580 ppm in one purple sapphire). In aV-Cr-Fe diagram (Fig. 3), the high-V (>90 ppm) suite mostlyplots along the V-Cr axis at Fe<10 %, but overall plots extend

up to Fe 65%. In contrast, the low-V blue sapphire suites (V<33 ppm, Cr<33 ppm) plot near the V-Fe axis at Fe 95–100 %.Both Mg and Ga ranges differ in the high-V (Mg 11–260, Ga7–110 ppm) and low-V (Mg 2–179, Ga 55–286 ppm) suites.Notably, V/Cr is >1 in 28 % of analyses and in 25 % of therubies V ratios range into very high values (V/Cr 26, V/Fe 78and V/Ti 97). BothMg andGa ranges differ in the high-V (Mg11–260, Ga 7–110 ppm) and low-V (Mg 2–179, Ga 55–286 ppm) suites. In considering genetic connections of corun-dum, the key element ratios including Cr, Mg, Fe, and Gacarry genetic implications for the metamorphic to magmaticcorundum suites. The first two elements are mostly higher inmetamorphic and metasomatic suites, whereas Fe and Ga aremostly higher in magmatic suites (Peucat et al. 2007;

Table 1 (continued)

Sample site Color Latitude Longitude No. V Cr Fe Ti Mg Ga V/Cr V+Cr/Ga

1.c 7 1 656 150 86 96.3 6.64 0.08

Kyauk-war (15)a Blue 22° 59′ 02″ 96° 32′ 30″ 1.r 1 1 4,913 57 27 69.9 0.56 0.03

1.c 1 <1.2 5,117 58 24 71.5 >0.48 0.03

Kyauk-war (15)a Blue 22° 59′ 02″ 96° 32′ 30″ 2.r 5 <0.8 3,717 124 67 74.2 >4.67 0.08

2.c 7 <1.1 3,853 207 122 96.7 >6.24 0.08

Ye-bauk-tayar (16)a Light blue 22° 55′ 40″ 96° 29′ 10″ 2.r 29 1 2,708 375 167 116.3 20.28 0.26

2.c 32 <1.1 2,633 297 179 116.3 27.83 0.28

Chaung-gyi (9)a Blue 22° 57′ 51″ 96° 30′ 32″ 1.r 6 14 1,165 104 65 103.7 0.4 0.19

1.c 5 9 1,126 111 57 102 0.56 0.14

Chaung-gyi (9)a Light blue 22° 57′ 51″ AM 1.r 18 <1.0 646 78 50 105.5 >18.0 0.18

1.c 18 <0.9 607 89 54 105.4 >20.2 0.18

Le-shu-kone-zan (17)a Dark blue 22° 58′ 30″ 96° 39′ 50″ 2.r 18 30 3,928 147 83 89.3 0.6 0.53

2.c 18 33 4,346 81 43 90.8 0.55 0.56

Pazaun-seik (18)a Dark blue 22° 57′ 43.3″ 96° 24′ 15.6″ 1.r 9 0.5 14,779 56 10 66.3 18 0.14

1.c 10 0.6 14,485 56 10 67.4 16.7 0.16

Pazaun-seik (18)a Dark blue 22° 57′ 43.3″ 96° 24′ 15.6′ 2.r 10 0.7 14,283 63 11 55.4 14.3 0.19

2.c 19 1.3 498 331 103 190 14.6 0.11

Lay-thar (19)a Medium blue 23° 00′ 26.3″ 96° 30′ 19.5″ 1.r 3 <0.6 4,804 52 27 125 >5.0 <0.03

1.c 3 <0.6 4,682 57 28 136 >5.0 <0.02

Lay-thar (19)a Pale blue 23° 00′ 26.3″ 96° 30′19.5″ 2.r 1 1 2,665 63 107 135 1 0.01

2.c 1 1 2,716 331 119 135 1 0.01

Average 8 sites 24 8 <4.2 3,933 139 63 96.4 >7.76 <0.14

Low V (<10 ppm) magmatic suites

Ye-bauk-tayar (16)a Blue 22° 55″40″ 96° 29″10″ 1.r 3 <1.1 1,171 24 23 230.5 >2.51 0.018

1.c 2 <1.1 1,557 24 19 223.3 >2.16 0.014

Shan-kone-zan (20)a Blue 22° 57′ 52″ 96° 29′ 56″ 1.r 4 <1.09 2,705 54 22 109.9 >4.25 0.046

1.c 4 <1.08 2,732 55 23 109.4 >4.80 0.046

Shan-kone-zan (20)a Blue 22° 57′ 52″ 96° 29′ 56″ 1.r 4 <1.0 3,927 29 40 110.2 >3.88 0.045

1.c 3 <1.1 3,808 30 38 111.9 >2.90 0.037

Le-shu-kone-zan (17)a Dark blue 22° 58′ 30″ 96° 39′ 50″ 2.r 9 6 2,420 81 8 284 1.5 0.05

2.c 9 8 1,952 71 19 286 1.13 0.06

Average 4 sites 8 4 <2.2 2,689 44 22 171 >1.82 <0.04

r rim, c corea Placerb Primary

Miner Deposita

Sutherland et al. 2009). As the Mogok suites have high Vvalues, we applied discrimination diagrams that used V+Cr/Ga and V/Cr in their parameters (see below). The high-VMogok field falls entirely within the metamorphic region of

trace element ratios in Fig. 7 and largely within the metamor-phic and transitional metamorphic regions (Ga/Mg<6), withonly minor overlap into magmatic values (Fig. 8). This isconsistent with a placer origin derived from the corundum-

Table 2 Oxygen isotope data of ruby and sapphire from the Mogok gemfield, Myanmar

Sample no. Mine name Mineral Color Sapphire (‰) Ruby (‰) Placer or primary

1 Sin-khwa Ruby Red 22.7 Placer

2 Lone-sho Sapphire Light blue 17.8 Placer

3 Pingu-taung Ruby Red 21.7 Primary

4 Htin-shu-taung Sapphire/Ruby Red 21.0 Placer

5 Ohnbin-ywe-htwet Sapphire/Ruby Yellow/red 22.7 22.9 Placer

6 Baw-mar Sapphire Red 24.2 Placer

7 Kadoke-tat Ruby Red 22.4 Primary

8 Mye-me Ruby Red 16.3 Placer

9 Chaung-gyi Sapphire Blue 15.6 Placer

10 Kolan Ruby Red 22.1 Primary

11 Baw-lone-lay Ruby Red 15.6 Primary

12 Baw-padan Ruby Red 21.5 Primary

13 Win-hta-yan Ruby Red 19.1 Primary

14 Pan-taw Sapphire Blue 16.6 Placer

15 Kyauk-war Sapphire Blue 16.7 Placer

16 Ye-bauk-tayar Sapphire 13.6 Placer

17 Le-shu-kone-zan Sapphire Blue 10.5 Primary

18 Pazaun-seik Sapphire Blue 12.1 Placer

19 Lay-thar Sapphire Blue 16.1 Primary

20 Shan-kone-zan Sapphire Blue 12.8 Placer

Fig. 3 V-Cr-Fe plot of ruby andsapphire from the Mogokgemfield, Myanmar

Miner Deposita

bearing marbles and skarns of the area. The intermediate low-V field overlaps metamorphic-magmatic ratios, suggestingmetasomatic origins, while the other low-V field falls wellwithin magmatic ratios. In comparison, metamorphic andmagmatic corundum from Pailin, Cambodia, basalt fieldsshow different average ratios to the Mogok corundums.

Latitude-longitude plots for Mogok area corundum showthe most extreme-V values (>1,000 ppm). They fall within alimited (10 km2) zone west of Mogok, where noted gem-bearing skarns concentrate. Samples beyond the wider high-V ruby zone (V 90–1,000 ppm), yield intermediate- to low-Vsapphires. The highest V-levels (>1,000 ppm) in the Mogoksuites cluster within in a limited (10 km2) NE-SW trendingzone (96.37–96.40° E, 22.90–22.9° N) (Fig. 4a), while pe-ripheral samples show moderate V (90–1,000 ppm) with low-Voutliers (<35 ppm).

Oxygen isotope composition

Oxygen isotope composition for the Mogok rubies range from15.6–24.2‰ and partly overlap a lower range for sapphires(10.6–22.7‰) (Table 2). Yui et al. (2008) previously reportedoxygen isotope composition of nine rubies from Mogok witha range of 21.6 to 25.7‰ (av. 22.9±1.5‰), which is similar tothe range of 18.9 to 22.0‰ for Mogok rubies reported byGiuliani et al. (2005). Such ranges typify carbonate-hosted,skarn and crustal magmatic/metasomatic corundum deposits(Giuliani et al. 2007; Garnier et al. 2008).

Plots of V against oxygen isotope composition for corun-dum at different site shows the higher V samples (V between1,000 to 5,500 ppm) range in δ18O from 16 to 23‰ (Fig. 4b).The intermediate-V corundum (V between 35 to 90 ppm)shows a lower δ18O range (12 to 18‰) and partly overlaps

Fig. 4 a Latitude-longitude plotof V-levels of ruby and sapphirewithin Mogok gemfield, Myan-mar. b V (ppm) vs. 18O of rubyand sapphire within Mogokgemfield, Myanmar. Sample lo-calities include Sk Sin-khwa, LsLone-sho, Pt’s Pingu-taung, HstHtin-shu-taung, Oby Ohnbin-ywe-htwet, Bm Baw-mar, KtKadoke-tat, MmMye-me, CgChaung-gyi, K Kolan, Bll Baw-lone-lay, Bp Baw-padan, WtyWin-hta-yan, Pt Pan-taw, KwKyauk-war, Ybt Ye-bauk-tayar,Lsk Le-shu-kone-zan, P Pazaun-seik, Lt Lay-thar, Skz Shan-kone-zan. The sample positions areshown relative to Mogok (filledtriangle)

Miner Deposita

into the low-V (V <35 ppm) magmatic corundum δ18O range(10 to 14‰). These O-isotope composition ranges suggestonly crustal metamorphic and magmatic processes generatedthe Mogok corundum suites, as they exceed those of mantleultramafic andmafic granulite values and largelymatch valuesfor skarns and marbles (Giuliani et al. 2007).

V-effects and ruby color

Ruby develops its red to orange pleochroismwhen Cr enters c.1 % of Al3+ sites. The Cr3+ introduces three unpaired orbitalelectrons (3d3) in a distorted octahedral ligand field, whichgives strong red color transmission (Nassau 2001). The coloris enhanced by red emission through UV-fluorescence, pro-viding Fe content is insufficient to bleed off energy from theCr. As Fe and Ti contents increase in ruby, electronic chargetransfer effects introduce purple and violet coloration whichgrades into sapphire ranges. In alpha-alumina, V substitutes inAl sites largely as V3+, whereas V4+ is minor. The V3+, as doesCr 3+, increases the sensitivity of corundum to X-ray colora-tion (Lambe and Kikuchi 1960). With V3+ substitution, the

simpler orbital structure (3d2) compared to Cr3+ produces greygreen to purple color effects at c. 3 % substitution. It givessharp absorption in the visible spectrum at 475 nm, which israrely reported in natural stones (Pryce and Runciman 1958;O’Donoghue 2006). Its presence in spectra relative to Ti lineshelps distinguish Mogok rubies from heat-treated Mong-Hsurubies (Themelis 2008).

Variations in V, Cr, and other chromophore values producestrong color zoning in some crystals. Trace elements in theMogok suites also vary in value from rim to core. In Pingu-tuang examples, LA-ICP-MS analysis shows that variable V(450–1,300 ppm) is enriched in cores, so that higher overall Cr(1,040–2,900 ppm) gives red rims (Table 1, Fig. 5). LA-ICP-MS imaging of the Pingu-taung ruby crystal also revealsmarked color and elemental zoning. The V levels either varywith Cr levels to give purple to mauve interiors passing intored rims and purple to grey interiors (Fig. 5) or remain fairlyuniform across the crystal relative to Cr giving deeper purpleinteriors. In Sin-khwa examples, LA-ICP-MS images revealthat high V (3,000–5,200 ppm) exceeds Cr (165–3,760 ppm),but a stronger crystal field effect of Cr3+ over V3+ enhances

3 mm

b c d e

f g h i

j k l m n

a

Fig. 5 Transmitted light and LA-ICP-MS trace element images of a rubycrystal from Pingu-taung,Mogok gemfield,Myanmar, showing color andelemental zoning from complex banded mauve core to irregular light redouter zone. a Imaged ruby crystal, bMg, c Ti, dV, eCr, f Fe, gGa, h Zr, i

Nb, j Sn, k Ce, l Ta,m Th, n U. Scale in ppm. Laser spot size is 60 μm.Note: vanadium is largely enriched along the red-colored rim togetherwith chromium. The Sn, Zr, and Ce enrichment of the top the left corner isa polishing artifact

Miner Deposita

red transmission over purple-grey (Nassau 2001), particularlyin Cr-enriched rims (Fig. 6). In Sin-khwa ruby, even with V≫Cr and V/Cr 12–26 red color and ruby status can be retained.The combined high-Vand Cr values in the outer zones gives apurple red color, while high-V with lower Cr in the inner zonesgives purple, mauve to grey colors. In Htin-shu-taung exam-ples, Cr (520–1,950 ppm) is≫V (150–465 ppm), but significantFe (900–1,230 ppm) shifts transmitted color into purple. SomeMye-me ruby samples have sufficient V (c. 1,400–1,600 ppm)over Cr (150–190 ppm) to modify colors into purple-pinks.High V, but dominant Cr may emphasize the purple hue andfluorescent effect prized in purple red Mogok rubies and needsdetailed evaluation. The elevated, but variable V levels noted in56 % of Mogok crystals in this study gives Van increased roleas a tracer in geographic typing of corundum suites.

Geographic typing

Both Mogok rubies (up to 0.5 wt % V, this study) and to somedegree the Mong-Hsu rubies (up to 0.1 wt %) (Mittermayr et al.2008) have higher V levels than those found in most rubieselsewhere (Muhlmeister et al. 1998). This characteristic may

assist in linking rubies to specific regional and sub-regionalareas. The highest V-levels (>1,000 ppm) in the Mogok suites(Fig. 4a), cluster in a limited (10 km2) NE-SW trending zone(96.37–96.40° E, 22.90–22.91° N), while peripheral samplesshow moderate V (90–1,000 ppm) with low-V outliers(<35 ppm). The highest V values correlate with ruby δ18Oranges between 16 and 22.5‰, within an overall 15.5–24.5‰range for high-V samples (Fig. 4b). The δ18O values for low-Vsapphires (<18‰) partly overlap the high-V δ18O range. Thesapphire samples with minimal V (<10 ppm) show the lowestδ18O (10–14‰). These results favor a geological control for theV- and O-isotope compositions found in the placer corundum.

Source of V and ruby genesis

The high-V, high δ18O ruby/sapphire samples occur in ruby-bearing marbles and skarns and nearby leucogranite intrusionswithin the Kyak-Pyat-That–Kyatpyin–Kathe region(Themelis 2008). The placer rubies suggest former marblehosts, as their δ18O values match those of marble-hostedrubies elsewhere (16–23‰) and mostly exceed typical values(8–16‰) for corundum in skarns (Garnier et al. 2006), apart

5 mm

a b c d e

f g h i

j k l m n

Fig. 6 Transmitted light and LA-ICP-MS trace element images ofa ruby crystal from Sin-Khwa, Mogok gemfield, Myanmar show-ing color and elemental zoning. a Imaged ruby crystal, b Mg, c

Ti, d V, e Cr, f Fe, g Ga, h Zr, i Nb, j Sn, k Ce, l Ta, m Th, nU. Scale in ppm. Laser spot size is 60 μm

Miner Deposita

from a primary-source ruby from Baw-lone-lay (15.6‰).Rubies in Asian carbonate-evaporite sequences are related tointernal processes during retrograde metamorphism at TPconditions between 620–670 °C and 2–4 GPa (Garnier et al.2008). Such an origin would imply that the high-V Mogokruby/sapphire suites stem from metasomatism of abnormallyV-enriched evaporatic components, rather than V introducedfrom fluids emanating from nearby leucogranites. This fitsknown ages for zircon-in-ruby dating (c. 32 Ma) (Khin Zawet al. 2008, 2012a, b, 2010) compared to leucocratic graniteintrusion ages (45 and 25Ma) (Kyaw Thu 2007). Pingu-tuangruby (δ18O 21.7‰), however, comes from an area of skarnsrelated to syenitic intrusions. Detailed comparisons of V andδ18O values of rubies hosted in Mogok marbles and skarns(Themelis 2008) are needed for more precise geneticinterpretation.

Discussion

Because many Mogok analyses show significant V and V/Cr>1, Cr + V values are used in chemical plots of these corun-dum suites. This modification in a V + Cr vs Fe + Ti plot,related to the corundum color, reinforces separations betweenhigh-and low-VMogok suites (Fig. 7a). AV + Cr/Ga vs Fe/Tidiagram clearly places the high-V suite within typical meta-morphic fields and the low-V suites in transitional to magmat-ic fields. Rubies and sapphires analyzed from the basalticPailin gemfield, Cambodia (Table 3) partially overlap theMogok fields (Fig. 7b). In contrast, V/Cr ratio plots only showlimited separation between Mogok suites, but clear separa-tions from Pailin suites (Fig. 8). Plots of Fe or Fe/Mg againstGa/Mg (Fig. 9) shows similar relationships to V/Cr vs. Ga/Mgdiagram, but the low-V Mogok field extends to higher Fe/Mg

Fig. 7 V, Cr and Fe, Tirelationships for ruby andsapphire, Mogok gemfield,Myanmar. a V+Cr vs Fe+Ti plot.Red squares (High-Vmetamorphic ruby/sapphiresuite), blue circles (Low-Vmetamorphic sapphire suite) andopen circles (Low-V magmaticsapphire suite). Large symbols areaverage data for the suites. Plotsfrom each site lie in enclosedareas. b V+Cr/Ga vs Fe/Ti plot.Red squares (High-V metamor-phic suite), blue circles (Low-Vmetamorphic suites), open circles(Low-V magmatic suite), closeddiamonds (High-V metamorphicsuite from Pailin, Cambodia) andopen diamonds (Low-V magmat-ic suite from Pailin, Cambodia),with large symbols representingaverage plots for each suite

Miner Deposita

than the high-V field due to high relative Fe contents. Thisdiagram helps in separating metamorphic sapphire and rubysuites (lower Fe, Ga/Mg) from magmatic sapphire suites(higher Fe, Ga/Mg) (Peucat et al. 2007; Sutherland andAbduriyim 2009; Sutherland et al. 2009). For Mogok suites,both the high-V and low-V suites extend from metamorphicinto transitional zones, whereas low-V sapphires with strongmagmatic affinity form a limited field. Mogok rubies plot wellapart from Pailin rubies (Fig. 8).

Harlow and Bender (2013) provide useful comparativeanalytical data from their Mogok ruby study. They collected

samples from field sites and some less controlled samplespurchased from miners, whereas this study only used welllocalized, field-collected samples. Harlow and Bender utilizedboth EMPA and LA-ICP-MS data sets and discussed varia-tions within their two trace element data sets from several sitesin the Mogok gem tract and farther afield in Myanmar. A fewof these sites also were sampled or nearby to sample sites inthe present study (Kadoke-tat, Kyauksin, Dattaw, Wetloo) butothers were farther afield in Myanmar than the sites sampledin the present study. Here, the LA-ICP-MS results are directlycompared, being the method used in this study. The

Table 3 LA-ICP-MS trace ele-ment data of ruby and sapphirefrom Pailin, Cambodia

Grain/spot V Cr Fe Ti Mg Ga V/Cr V+Cr/Ga

Pink/red group

J1.1 core 30 1,332 4,523 339 279 38 0.023 35.84

J1.1 rim 29 1,328 4,488 338 275 29 0.022 46.79

J1.2 core 29 1,218 4,726 301 273 34 0.024 36.68

J1.2 rim 31 1,716 4,685 307 257 35 0.018 49.91

J1.3 core 36 1,010 4,724 395 312 41 0.036 25.51

J1.3 rim 41 1,830 4,597 390 299 42 0.022 44.55

J1.4 core 25 1,219 3,624 141 239 26 0.021 47.85

J1.4 rim 25 1,261 3,614 145 237 25 0.020 51.44

J1.5 core 28 2,198 4,384 331 280 37 0.013 60.16

J1.5 rim 33 1,991 4,410 311 261 31 0.017 65.29

Average (10) 31 1,509 4,378 300 271 34 0.021 45.29

Blue group

J6.1 rim 13 1.09 6,637 953 19 218 11.9 0.065

J6.1 core 11 1.05 2,416 100 1 157 10.5 0.077

J5.1 rim 11 1.14 6,623 544 20 217 9.6 0.101

J5.1 core 7 0.41 1,883 65 2 94 17.1 0.008

Average (4) 11 0.92 4,390 416 11 172 12.3 0.069

Fig. 8 V/Cr vs. Ga/Mg plot ofruby and sapphire from theMogok gemfield, Myanmar. Redsquares (High-V metamorphicsuite), filled blue circles (Low-Vmetamorphic suite), open circles(Low-V magmatic suite). ThePailin, Cambodia fields (P) in-clude closed diamonds (meta-morphic ruby/sapphire suite) andopen diamonds (magmatic sap-phire suite). The large symbolsrepresent average plots for eachsuite. Ga/Mg ratio 3–6 representstransition zone between meta-morphic and magmatic blue sap-phire suites defined by Peucatet al. (2007)

Miner Deposita

comparisons of Vand relevant V, Cr ratios are shown Table 4.Plotting comparative data on geochemical variation figureswas not feasible, as Harlow and Bender mostly list Fe and Mgas below detection limit (bdl). In the present study, bdl (ppm)for Fe (av. 8.1) and Mg (av. 0.8) are much lower than inHarlow and Bender (Fe 544–2,332; Mg 6–151). For theirEMPA results, Fe and Mg similarly range down into bdlvalues (dl Fe 60–106, Mg 22–70 ppm). Table 4 herein clearlyillustrates higher V and V/Cr (n=46), with V/Cr>1 (n=13)relative to Harlow and Bender LA-ICP-MS data (n=52), withV/Cr>1 (n=8). Their EMPA high content for V (2,634 ppm)is half the LA-ICP-MS high data found in this study(5,213 ppm). We note their V-rich ruby locality (KodokeTat) lies close to the high-V (>1,000 ppm) ruby zone detectedin this study (Fig. 4a). The denser sampling in the presentstudy revealed an unsuspected high-V ruby zone within theMogok region, although circumscribed in geographic extent.

Although the LA-ICP-MS ruby results make the best com-parison in the two studies, observations on the wider Harlowand Bender data set, including EMPA results, can be consid-ered here. Their Mogok EMPA data (eight sites) give Crvalues, ranging between detection limit to 25,000 ppm, andindicate high-Cr rubies (av.>1,000 ppm) exist at Dattaw,Wetloo, Kadoke-tat, Saw Baw, Namya, and Mongshu, with

the most extreme values at Namya and Mongshu. The Vvalues range between 50 to 2,635 ppm, with V-rich rubies(>90 ppm) present at most sites, but only in rare extreme-Vrubies (av. >1,000 ppm) are recorded at Kodoke Tat andMongshu. The EMPA values for Cr are generally 1.3 to 5times, and V 1.2 to 2.7 times greater than the respective LA-ICP-MS values. Such differences from the two analyticalmethods are similar to those noted in comparative EDXRFand LA-ICP-MS analyses on Mogok sapphires (Kan-Nyuntet al. 2013). They were related to factors such as size ofmeasurement area (smaller for LA-ICP-MS), detection limits(lower for LA-ICP-MS) and influences of inclusions. Thus,even in the wider analytical context, the rubies in this studyemphasize the exceptional V values within the designatedenrichment zone.

Harlow and Bender (2013) LA-ICP-MS results did not listSi values due to a high detection limit (1,400–6,550 ppm), buttheir EMPA data included Si, which ranged from 80 to6,080 ppm. They considered levels of 150–500 ppm Si couldbe accommodated in the corundum lattice, but that higherlevels probably represent fine-scale inclusions. Ruby LA-ICP-MS analyses in the present study gave Si ranges from70 to 305 ppm, values easily substituted into the structure. TheHarlow and Bender Ga values differ significantly in their

Fig. 9 Plot of Fe or Fe/Mgagainst Ga/Mg showing similarrelationships to V/Cr vs. Ga/Mgdiagram (Fig. 8), but the low-VMogok field extends to higher Fe/Mg than the high-V field due togreater relative Fe contents. Thisdiagram helps in separatingmetamorphic sapphire and rubysuites (lower Fe, Ga/Mg) frommagmatic sapphire suites (higherFe, Ga/Mg) as indicated byPeucat et al. (2007), Sutherlandand Abduriyim (2009), Suther-land et al. (2009)

Table 4 Ruby LA-ICP-MS element comparisons, this study and Harlow and Bender (2013)

Study area Sites analysis V range, ppm av. V/Cr range av. V+Cr range av. V/(V+Cr) av. Ga/Mg av.

This study 13 92–5,044 0.03–25.9 110–7,718 0.03–0.96 0.13–12.33

Mogok 46 759 2.35 1,827 0.77 2.14

Harlow and Bender 4 64–1,472 0.08–5.34 169–3,209 0.08–0.84 0.05–0.46a

Mogok 24 354 1.12 1,086 0.38 0.17

Harlow and Bender 7 64–1,472 0.03–5.34 295–7,421 0.06–0.84 0.05–0.42b

Myanmar 52 286 0.63 1,671 0.27 0.25

a Kadoke-tat only, n=2b Sagyin only, n=5

Miner Deposita

EMPA and LA-ICP-MS data sets and they considered the LA-ICP-MS values the more accurate. This is important in thegenetic context, as higher Ga would shift the metamorphicruby Ga/Mg ratios towards more transitional magmatic values(Peucat et al. 2007). In their study, Harlow and Bender (2013)sought to clearly identify skarn-origin trace element “finger-prints” in their ruby analyses. They did not fully identify suchpatterns, but thought high Si may mark a pointer. Khin Zawet al. (2012a, b) also recorded potential skarn trace elementenrichments in both V-rich ruby and sapphire. This involvedsignificant values (ppm) of B (11 to 82), Si (1,600 to 4,900),Ca (1,800 to 4,600), Ga (280 to 800), and Sn (2 to 35).

Conclusions

This geochemical study on alluvial gem corundum deposits inthe Mogok area shows that V plays a greater role as a rubytracer than is generally considered. Significantly higher V(1,000 to 5,500 ppm) characterized ruby deposits in a con-fined zone west of Mogok, suggesting a local source control.The high-V ruby data provides added potential in explorationfor and in geographic typing of ruby sources. High V intandem with Cr in rubies can modify the chromophoric col-oration and gem quality.

Acknowledgments Ross Pogson, Geoscience, Australian Museum,helped with geochemical plotting programs. Sang Ding helped in LA-ICP-MS analysis of the rubies and sapphires. This study is supported byCODES ARCCentre of Excellence, University of Tasmania. The authorsare deeply indebted to reviewers and Georges Beaudoin, Editor-in-Chief,Mineralium Deposita for his insightful review, constructive comments,and suggestions to the substantial improvement of this manuscript.

References

Barley ME, Pickard AL, Khin Zaw, Rak P, Doyle MG (2003) Jurassic toMiocene magmatism and metamorphism in the Mogok metamor-phic belt and the India-Eurasia collision in Myanmar. Tectonics 22:1019. doi:10.1029/2002tc001398

Bender F (1983) Geology of Burma. Borntraeger, Berlin, p 293Bertrand G, Rangin C, Maluski H, The GIAC Scientific Party (2001)

Diachronous cooling along the Mogok Metamorphic Belt (ShanScarp, Myanmar): the trace of the northward migration of theIndian syntaxis. J Asian Earth Sci 19:649–659

Clegg ELG (1941) The Cretaceous and associated rocks of Burma. GeolSurv India Mem 74:101

Garnier V, Maluski H, Giuliani G, Ohnenstetter D, Schwarz D (2006) Ar-Ar and U-Pb ages of marble-hosted ruby deposits from central andsoutheast Asia. Can J Earth Sci 43:509–532. doi:10.1139/e06-005

Garnier V, Giuliani G, Ohnenstetter D, Fallick AE, Dubessy J, Banks D,Vinh HQ, Lhomme T, Maluski H, Pêcher A, Bakhsh KA, Long PV,Trinh PT, Schwarz D (2008) Marble-hosted ruby deposits fromCentral and Southeast Asia: towards a new genetic model. OreGeol Rev 34:169–191. doi:10.1016/j.oregeorev.2008.03.003

Giuliani G, Fallick AE, Garnier V, France-Lanord C, Ohnenstetter D,Schwarz D (2005) Oxygen isotope composition as a tracer for theorigins of rubies and sapphires. Geology 33:249–252

Giuliani G, Ohnenstetter D, Garnier V, Fallick AE, Rakotondrazafy M,Schwarz D (eds) (2007) The geology and genesis of gem corundumdeposits. Mineralogical Association of Canada, Québec

Graham I, Sutherland L, Khin Zaw, Nechaev V, Khanchuk A (2008)Advances in our understanding of the gem corundum deposits of theWest Pacific continental margins intraplate basaltic fields. Ore GeolRev 34:200–215. doi:10.1016/j.oregeorev.2008.04.006

Harlow GE, Bender W (2013) A study of ruby (corundum) compositionsfrom the Mogok Belt: searching for chemical fringerprints. AmMineral 98:1120–1132

Hughes RW (1997) Ruby and sapphires. KWH, BoulderIyer LAN (1953) Geology and gemstones of the Mogok stone tract,

Burma. Geol Surv India Mem 82:129Kammerling RC, Scarratt K, Bosshart G, Jobbins EA, Kane RE, Gubelin

EJ, Levinson AA (1994) Myanmar and its gems—an update. JGemmol 24:3–40

Kan-Nyunt H-P, Karamppeles S, Link K, Kyaw Thu, Kiefert L, Hardy P(2013) Blue sapphires from the Baw Mar Mine in Mogok. GemsGemology 49(4):223–232

Khin Zaw (1989) Comments on transcurrent movements in theMyanmar-Andaman sea region. Geology 17:93–95

Khin Zaw (1990) Geological, petrogical and geochemical characteristicsof granitoid rocks in Burma: with special reference to the associatedW-Sn mineralization and their tectonic setting. J SE Asian Earth Sci4:293–335. doi:10.1016/0743-9547(90)90004-w

Khin Zaw (1998) Geological evolution of selected granitic pegmatites inMyanmar (Burma): constraints from regional setting, lithology, andfluid-inclusion studies. Int Geol Rev 40:647–662. doi:10.1080/00206819809465229

Khin Zaw, Sutherland FL, Graham IT, McGee B (2008) Dating zirconinclusions in gem corundums from placer deposits, as a guide totheir origin. 33rd International Geological Congress (IGC), 5–14August, 2008, Oslo, Norway, CD-ROM

Khin Zaw, Sutherland FL, Graham IT, Meffre S, Thu K (2010) Datingzircon inclusions in gem corundum deposits and genetic implica-tions. 13th Quadrennial IAGOD Symposium. Adelaide, SouthAustralia

Khin Zaw, Sutherland FL, Meffre S, Yui T-F, Thu K (2012a) Distinctivegeochemistry among placer ruby suites, Mogok gemfield,Myanmar. 34th International Geological Congress (IGC), 5–10August, 2012, Brisbane, Australia, CD-ROM

Khin Zaw, Sutherland FL, Meffre S, Yui TF, Thu K (2012b) Distinctivegeochemistry among placer ruby suites, Mogok gemfield,Myanmar. 34th International Congress, Brisbane, Australia, 5–10th August 2012, Abstracts (CD), [3243]

Khin Zaw,Meffre S, Lai CK, SantoshM, Burrett FC, Graham IT,ManakaT, Salam A, Kamvong T, Cromie P (2014) Tectonics andmetallogeny of mainland SE Asia—a review and contribution.Special issue on Tectonics and Metallogeny of mainland SE Asia.Gondwana Res 26:5–30. doi:10.1016/j.gr.2013.10.010

Kyaw Thu (2007) The igneous rocks of the Mogok Stone Tract: theirdistribution, petrography, petrochemistry, sequence, geochronologyand economic geology. Yangon University, Yangon

Lambe J, Kikuchi C (1960) Spin resonance of V2+, V3+, V4+ in a-Al2O3.Phys Rev 118:71–77

Large RR, Danyushevsky L, Hollit C, Maslennikov V, Meffre S, GilbertS, Bull S, Scott R, Embsbo P, Thomas H, Singh B, Foster J (2009)Gold and trace element zonation in pyrite using a laser imagingtechnique: implications for the timing of gold in orogenic andCarlin-style sediment-hosted deposits. Econ Geol 104:635–668

Mitchell AHG, Ausa CA, Deiparine L, Hlaing T, Htay N, Khine A (2004)The Modi Taung–Nankwe gold district, Slate belt, central Myanmar:mesothermal veins in aMesozoic orogen. J Asian Earth Sci 23:321–341

Miner Deposita

Mitchell AHG, Htay MT, Htun KM, Win MN, Oo T, Hlaing T (2007)Rock relationships in the Mogok metamorphic belt, Takon toMandalay, central Myanmar. J Asian Earth Sci 29:891–910

Mittermayr F, Konzett J, Hauzenberger C, Kaindl R, Schmiderer A(2008) Trace element distribution, solid-and fluid inclusions inuntreated Mong Hsu rubies Geophysical Research Abstracts, p1076

Muhlmeister S, Fritsch E, Shigley JE, Devourd B, Laurs BM (1998)Separating natural and synthetic rubies on the basis of trace elementchemistry. Gems Gemology 34:80–101

Nassau K (2001) The physics and chemistry of colour, 2nd edn. Wiley,New York

O’Donoghue M (2006) Gems, 6th edn. Butterworth-Heinemann, OxfordPeucat JJ, Ruffault P, Fritsch E, Bouhnik-Le Coz M, Simonet C, Lasnier

B (2007) Ga/Mg ratio as a new geochemical tool to differentiatemagmatic from metamorphic blue sapphires. Lithos 98:261–274.doi:10.1016/j.lithos.2007.05.001

Pryce MHL, Runciman WA (1958) The absorption spectrum of corun-dum. Discuss Faraday Soc 26:34–42

Searle DL, Haq BT (1964) TheMogok Belt of Burma and its relationshipto the Himalayan Oronogeny. In: Proc, 24th Int. Geol. Congr., PartII, pp 132–161

Searle MP, Noble SR, Cottle JM, Waters DJ, Mitchell AHG, Hlaing T,Horstwood MSA (2007) Tectonic evolution of the Mogok meta-morphic belt, Burma (Myanmar) constrained by U-Th-Pb dating ofmetamorphic and magmatic rocks. Tectonics 26, TC3014. doi:10.1029/2006tc002083

Sutherland FL, Abduriyim A (2009) Geographic typing of gem corun-dum: a test case from Australia. J Gemmol 31:203–210

Sutherland FL, Khin Zaw, Meffre S, Giuliani G, Fallick AE, Graham IT,Webb GB (2009) Gem-corundum megacrysts from east Australianbasalt fields: trace elements, oxygen isotopes and origins. Aust JEarth Sci 56:1003–1022. doi:10.1080/08120090903112109

Themelis T (2008) Gems & mines of Mogok. A & T, BangkokValley JW, Matthew NM, Kohn J, Niendorf CR, Spicuzza MJ (1995)

WG-2, a garnet standard for oxygen isotope ratios: strategies forhigh precision and accuracy with laser heating. GeochimCosmochim Acta 59:5223–5231

Waltham T (1999) The Ruby mines of Mogok. Geol Today 15:143–149Yui TF (2000) Preliminary results on CO2 laser-fluorination system for

O-isotope micro-analysis of silicate/oxide grain separates. J GeolSoc China 43:237–246

Yui TF, Khin Zaw, Wu CM (2008) A preliminary stable isotope study onMogok ruby, Myanmar. Ore Geol Rev 34:192–199

Miner Deposita