Sa

La

b

c

d

a

ARRA

KDSRPSUN

1

p4b0gmasedatP

2

1d

Spectrochimica Acta Part A 80 (2011) 112– 118

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular andBiomolecular Spectroscopy

jou rn al hom epa ge: www.elsev ier .com/ locate /saa

pectroscopic research on ultrahigh pressure (UHP) macrodiamond at Copetonnd Bingara NSW, Eastern Australia

. Barrona,d,∗, T.P. Mernaghb, B.J. Barronc,d, R. Pogsona

The Australian Museum, 6 College Street, Sydney, NSW 2010, AustraliaGeoscience Australia, GPO Box 378, Canberra, ACT 2601, AustraliaConsulting Petrologist, 7 Fairview Avenue, St Ives, NSW 2075, AustraliaSchool of Biological and Environmental Sciences, University of New South Wales, Australia

r t i c l e i n f o

rticle history:eceived 3 August 2010eceived in revised form 16 February 2011ccepted 2 March 2011

eywords:iamondpectroscopyaman, infrared, X-ray, UV-fluorescenthanerozoicubduction

a b s t r a c t

Millions of macrodiamonds were mined from Cenozoic placers across Eastern Australia, 98% from withinthe Copeton and Bingara area (85 km across) in the Phanerozoic New England region of New South Wales(NSW). Raman spectroscopy of inclusions in uncut diamond, from the Copeton and Bingara parcels,identifies them as ultrahigh pressure (UHP) macrodiamond formed during termination of subduction bycontinental collision. Infrared spectral properties of the two parcels are critically similar in terms of nitro-gen abundance (low in zoned diamond, high in unzoned diamond), requiring a pair of different growthmechanisms/protoliths. Within each parcel, the degrees of nitrogen aggregation are relatively strong andcoherent, but they are so different from each other (moderate aggregation for Bingara, strong for Copeton)that the two parcels require separate primary and local sources. The local sources are post-tectonic alkalibasaltic intrusions which captured UHP minerals (garnet, pyroxene, diamond) from eclogite-dominated

ltrahigh pressure UHP aggregation

UHP terranes (density stranded at depth—mantle, lower crust). X-ray diffraction studies on Copetondiamond indicate a normal density, despite previous reports of anomalously high density. For non-fluorescent diamond, a 2nd order Raman peak, which is prominent in theoretical perfect diamond andin African cratonic diamond, is suppressed in Copeton and Bingara UHP macrodiamond. Pervasive defor-mation during macrodiamond growth probably causes this suppression, the strong nitrogen aggregation,and the exceptional durability documented through industrial use.

. Introduction

About a million carats of macrodiamonds (average 4 stoneser carat) were mined from placers and paleoplacers within a00 × 3000 km belt in Eastern Australia [1], see Fig. 1. Despite thisroad distribution, over 98% of production was confined to just.6% of this area: within the adjacent districts of Copeton and Bin-ara in the New England region of New South Wales (NSW). In theined paleoplacers, typical diamond-related minerals are virtually

bsent, hampering efforts to trace the diamond back to hardrockources. The eastern margin of Australia experienced recurrentpisodes of prolonged westerly subduction, and termination of sub-uction due to collision with microcontinents, causing accretion

nd growth of Eastern Australia throughout the period late Neopro-erozoic to mid Triassic [2], with most prominent episodes being ofhanerozoic age.

∗ Corresponding author at: The Australian Museum, 6 College Street, Sydney, NSW010, Australia. Tel.: +61 2 94495839; fax: +61 2 94495839.

E-mail address: barronjandl@optusnet.com.au (L. Barron).

386-1425/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rioi:10.1016/j.saa.2011.03.003

Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.

Previous research on these macrodiamonds has demonstratedits unique subduction-related non-cratonic features [3–16] andtwo source models have been proposed: (A) diamonds were pre-existing in volumes of metastably cool lithosphere (ancient orPhanerozoic) emplaced under Australia and (B) diamonds formedduring subduction, either ancient or Phanerozoic. Raman spec-tral determinations, of inclusion internal remnant pressures (Pr) inuncut Copeton diamond, set up an PT array of inclusion-entrapmentloci. The intersection of these restricts the diamond formation con-ditions to 47 ± 4 kbar and 250–840 ◦C, identifying them as ultrahighpressure (UHP) macrodiamond [9]. These results are consistentwith an approximate thermodynamic PT locus derived from com-positions of garnet and clinopyroxene inclusions [12].

Our preferred tectonic model for this UHP macrodiamond hasit formed during various continental collision events which ter-minated repeated episodes of Phanerozoic “young” subduction.The diamond is characterized by strong pervasive deforma-

tion during growth, which has strongly accelerated nitrogenaggregation [12,14], thereby mimicking prolonged storage. Com-pared with cratonic diamond, anomalously high Pr values onolivine/garnet/pyroxene inclusions are characteristic of Copeton

ghts reserved.

L. Barron et al. / Spectrochimica Ac

Fb(

mltiuetmrsata

sai

ig. 1. Palaeoalluvial and alluvial diamond occurrences, and relevant geologicaloundaries within Eastern Australia. Copeton is east of Bingara, whereas WalchaW) is about 100 km south of Copeton.

acrodiamond [9], and inherently constrain diamond formation toow temperature UHP conditions [9]. These higher Pr values excludehe possibility of plastic release of internal pressures on inclusionsn diamond during delivery to the earth’s surface, unlike the contin-ation of plastic behaviour in host garnet during exhumation of thexposed UHP terrane at Kokchetav [17]. Our diamond surface fea-ures indicate magmatic delivery to the earth’s surface [9,12–14],

ost likely by post-tectonic alkali basaltic intrusions which carryelatively abundant UHP minerals (garnet, pyroxene, for analy-es see [18]), captured from high density UHP terranes strandedt depth (upper mantle, lower crust) [18]. Despite this UHP link,here are no exposed UHP terranes in NSW and microdiamonds arebsent.

Copeton/Bingara macrodiamonds are characterized by roundedurfaces (see Fig. 2) with more than 90% of shiny stones havingbundant to overlapping microdiscs. The most common inclusions coesite (Fig. 2d), having an equant to tapering prismatic to platy

ta Part A 80 (2011) 112– 118 113

(Fig. 2d) shape, the latter two with imposed six-sided symme-try. Important physical structures include euhedral indentations(roughly 5% of stones), pitted depressions (15%, see Fig. 2d) andnaats (30%, a narrow belt with multiple twins).

Our overall research on more than 250 uncut Eastern Australianmacrodiamonds showed that 97% of inclusions tested failed togive a Raman response, so the Raman inclusion technique is notan efficient way for testing random uncut macrodiamond as non-cratonic in origin. Menneken [19] compiled published details aboutRaman spectra of microdiamond from well-defined non-cratonicsources (exhumed UHP terrane; Hawaiian mantle-circulation; backarc basin in Japan), and proposed a test for identifying non-cratonicdiamond: a 2nd order Raman peak prominent in cratonic diamondis absent from non-cratonic diamond. First, the Menneken test istrialed and calibrated on African macrodiamond, then it is appliedto alluvial macrodiamond from Copeton, Bingara and Walcha (allin New England NSW), Argyle (West Australia) and Cempaka (Kali-mantan, Indonesia).

We make the first detailed comparison of nitrogen abundancesand aggregation values in diamond from Copeton [10] and Bingara[12,14,16], and use this to demonstrate that there must be multiplelocal intrusions capturing diamonds from different formation rocks(at depth). Some Copeton diamond was X-rayed to obtain the theo-retical density, to compare with previous reports of NSW diamondhaving anomalously high specific gravity [1].

2. Experimental

Published nitrogen abundance and aggregation results for dia-mond from Copeton and Bingara were derived from infraredspectroscopy using established procedures [10,12,14,16].

The Raman equipment comprises a Dilor SuperLabram spec-trometer, with a holographic notch filter (600 and 1800 g/mmgratings), liquid nitrogen-cooled 2000 pixel CCD detector, and a514.5 nm Melles Griot 543 argon ion laser (5 mW at the sample).The spectral resolution was set at 2 cm−1 (slit width of 100 �m).The microscope uses a 50× ULWD Olympus microscope objective,focusing the laser spot to 2 �m in diameter and 5 �m deep.

A PanAlytical X’Pert Pro X-ray Diffractometer with a �–� geo-niometer was used to X-ray powdered diamond, operating at40 mA, 45 kV of Cu K� radiation, with gas-filled proportionalcounter, graphite monochromator, 1◦ divergence slit, 2◦ anti-scatter slit and 0.1 mm receiver slit.

3. Theory

3.1. Nitrogen abundances and aggregation states

Nitrogen is the most abundant element that substitutes for car-bon atoms in the diamond structure, with values up to 3300 ppmreported for macrodiamond [20,21]. Reviews [6,22] of nitrogenaggregation in diamond show N starts off as singly substitutedfor carbon on random lattice sites, then these rapidly aggregateto place N atoms on two adjacent sites (called IaA). By a slowprocess, IaA aggregates to place N atoms on four adjacent sites(called IaB). Higher N abundances, longer storage times, higherstorage temperature, and strong plastic deformation during/aftergrowth all increase the aggregation state, see Fig. 3. The dashedlines (Fig. 3) are isomaturity trends for nitrogen aggregation, cal-culated as a combination of storage time and temperature (see[23] for equations and coefficients, noting that the results are in“time equivalents” etc. since deformation is ignored in the calcu-

lation). A significant to major fraction of singly substituted N hasbeen reported in microdiamond from exhumed UHP terranes [24],but also in some kimberlitic diamond from cratonic terranes [6],indicating a short period of confinement (1–20 MY) between for-

114 L. Barron et al. / Spectrochimica Acta Part A 80 (2011) 112– 118

Fig. 2. Plane light photographs of Copeton and Bingara paleoalluvial macrodiamond. (a) Streak of Luck prospect (Copeton) showing rounded dodecahedral shapes, manywith naats. (b) White 0.43 ct 4 mm stone from Mount Ross (Copeton) showing euhedral indents on apices of a rounded octahedral shape, with a naat cutting the centrali ent ofo 1 ct stT shape

msld

FCl

ndent. Diamond 96.10, photo by David Barnes courtesy of NSW Minerals (Departmf base of indent. The naat changes the indent slope. (d) Closeup of pale yellow 0.2he top inclusion is a steeply dipping plate that is 20 �m thick, and has a six-sided

ation and delivery to the earth’s surface. If a diamond has no singlyubstituted nitrogen, {N} = {IaA} + {IaB} for {concentrations}, and aonger storage time “equivalent” is inferred. In a study of 59 cratoniciamonds from a single xenolith, values of nitrogen abundance and

ig. 3. Nitrogen abundance and aggregation state (IaA to IaB) for diamond fromopeton C1 C2 (heavy lines) [10] and Bingara B1 B2 (light lines) [12,14,16]. Dashed

ines are isomaturity trends for nitrogen aggregation.

Industry and Investment). (c) Close up of centrally located indent on (b), with insetone from the Monte Cristo prospect (Bingara) showing three inclusions of coesite.

imposed by the host.

aggregation were determined to be from a single growth eventand storage regime [23]—from this data, we calculate the varia-tion (� = 0.23 in log years storage), and use this hereafter as thenull hypothesis for this above special case.

3.2. Raman

The Menneken test (using 2nd order Raman spectra) was tri-aled on uncut African cratonic macrodiamond to determine thenature and degree of compliance. The critical peak is in the range2450–2500 cm−1, see Fig. 4 for the procedure of extracting valuesfor the Raman background and the S/N ratio of this critical peak.The spectral count is accumulated (10×) with a count time in therange of 12–60 s (longer count times used for noisy backgrounds).Multiple determinations using different count times confirm theS/N ratio can be measured reliably. Modelling in theoretical per-fect diamond [25,26] shows the critical peak is due to two-phononinteractions on non-adjacent lattice sites.

3.3. XRD

Mined from the Streak of Luck paleoplacer (Copeton), six small

white clear macrodiamonds (non-fluorescent by eye to UV light, noinclusions) were crushed to 0.6 ct of powder and X-rayed. The pow-der was scanned from 40◦ to 95◦ of 2�, counting 1 s per 0.02◦ 2� step.Only the peaks with Miller indices [27] designated as <2 2 0> and

L. Barron et al. / Spectrochimica Ac

Fig. 4. Determination of Signal/Noise ratio (S/N) of critical 2nd order Raman peakin non-fluorescent African diamond D16213.02. The arbitrary units (a.u.) representthe total counts over the count interval (15 s in this case), accumulated ten times.First the baseline T-T is drawn from the shoulder of the small peak at 2200/cmacross to the base of the spectra beyond 2690/cm. A vertical line from the criticalpeak (2460/cm) is drawn, intersecting the T-T line. The baseline value for the criticalpeak is the intensity at B (450 a.u.), whereas the strength of the critical peak is thelength of the dashed line up from S (1020–450 = 570 a.u.). Along a relatively straightportion of the spectral trace, two parallel lines enclosing the spectral “noise” aredFc

<aF2wtofahl(l

4

4m

imsnobsrptottoir

rawn, yielding the short heavy lines at 2N as twice the noise (2N = 55, N = 22.5 a.u.).rom these intermediate results, the Raman background is 450/15 = 30 cps, and theritical peak ratio S/N = 570/22.5 = 25.3.

3 1 1> were selected for cubic unit cell edge calculations, using 2�ngles derived by the supporting PanAlytical HighScore software.or <2 2 0> and <3 1 1 > peaks respectively, the Cu K�-1 and Cu K�-

peaks were treated separately to obtain four d-spacings, whichere converted to a unit cell size (implying a theoretical density) by

he equation ao = d√

(h2 + k2 + l2) where h k l are the Miller indicesf the chosen peaks [27, pp. 261–327]. Macrodiamond densitiesrom many countries are reported [19], and fall in two groups:

lower density group of 11 stones (3.512 ± 0.0022 g/cm3), and aigher density group of 15 stones (3.5153 ± 0.00017 g/cm3), estab-

ishing a density reference. Previous reported values of density3.567–3.578 g/cm3) for Bingara diamond [1] are clearly anoma-ous.

. Results and discussion

.1. Interpretation of some physical structures on UHPacrodiamond from NSW

The following physical features, listed in order of decreas-ng frequency of occurrence, are characteristic of Copeton/Bingara

acrodiamond: rounded crystals, microdiscs, naats, pitted depres-ions and euhedral indentations. The shiny cylindrically curvedature of the euhedral indents in Fig. 2b indicates simultane-us but stable growth of both touching minerals. The interactionetween the naat and the euhedral indentation (Fig. 2b and c—i.e.lope-change and not offset) indicates the naat is a growth featureather than a post-crystal structure. Single stones may have multi-le naats, both with parallel and crossing relations [13] suggestinghat plastic deformation is involved. There are abundant microdiscsn the shiny indents (Fig. 2b and c) and on the outer surface ofhis stone, indicating that the shiny surface and microdisc forma-

ion may involve only a trace change of the surface and volumef the diamond, perhaps caused by hot magmatic fluids percolat-ng along grain boundaries in a xenolith. The formation of suchounded shapes for macrodiamond is typically stated to be due

ta Part A 80 (2011) 112– 118 115

to strong resorption with only 55–1% volume preservation [13].However, the conceptual act of progressively adding thin outershells of diamond to the stone in Fig. 2b shows that the indentsmust rapidly increase in size until they reach each other with theaddition of only 25–30% diamond. Conversely, this small changemust represent the maximum degree of resorption of this partic-ular well-rounded stone. Thus the present rounded shapes/sizesof some Copeton/Bingara diamond may be close to their originalgrowth shapes/sizes (70-75% preservation).

Distinct from the euhedral indentations, the strongly pitteddepressions are somewhat rounded and fairly shallow. They oftenoccur in groups (Fig. 2d, in a belt parallel to but offset from an edge).They are not etched equivalents of the euhedral indents, becausethey are too shallow, and do not have similar crystallographic prop-erties (symmetry, apical location). In groups, their crystallographicorientation is apparently related to the intersection of the surfacewith dislocations in the crystal [13], so they may occur throughmagmatic corrosion, localized at points of weakness. Their etchroughness is significantly less than that shown on the base of euhe-dral indentations (Fig. 2b and c).

4.2. Nitrogen maturation in diamond based on content andaggregation

See Fig. 3 for nitrogen characteristics of diamond from Copeton[10] and Bingara ([14] with label-error corrected [16] using[12]). For Eastern Australian macrodiamond in general, andCopeton/Bingara UHP diamond in particular [10,14], there is nosingly substituted nitrogen left and the nitrogen is aggregated ontoadjacent pair IaA and quartet IaB lattice sites. Based on the actualpoint data, the spread for each of these data fields is no more thanthe null hypothesis for a simple growth/storage regime (see Section3.1), so Fig. 3 results are simple, coherent and exclude complexi-ties that have been reported for some diamond suites (for example[28]).

Diamond types C1 and B1 (Fig. 3) have low N levels in com-plexly zoned white stones, whereas C2 and B2 have high N levelsin unzoned yellow stones. These differences imply two entirelydifferent growth mechanisms in an associated pair of primary pro-toliths. Cathodoluminescence and X-ray topographic studies showboth zoned and unzoned Copeton/Bingara diamond were stronglyand pervasively deformed (brittle, plastic) during growth [10,14].Despite these features in common, the fields for Copeton and Bin-gara diamond in Fig. 3 show almost no overlap. This means thatthe Copeton district did not supply alluvial diamond to the adja-cent Bingera district or vice versa, requiring multiple local sourcesfor diamond, tapping two separate UHP source terranes trapped atdepth. These source UHP terranes not only differ in terms of rela-tive diamond nitrogen aggregation, the diamonds have moderatelydifferent maximum N abundances (2000 ppm C2; 2700 ppm B2)and different splits (C1/C2 at 220 ppm N and no overlap; B1/B2 at500 ppm N with trace overlap).

Copeton diamond has significantly stronger relative nitrogenaggregation than Bingara diamond. Since the relatively highernitrogen abundances in Bingara diamond should promote strongeraggregation, this actual converse behaviour indicates the defor-mation/storage conditions for Copeton diamond must have beencrucially more extreme than for Bingara diamond. The actual stor-age time for NSW Phanerozoic UHP macrodiamond is expectedto be considerably less than for ancient cratonic diamond, yetCopeton/Bingara diamond has N aggregation states similar to thosereported for cratonic diamond [21], emphasizing the extreme

influence of pervasive deformation during the growth of thismacrodiamond. N aggregation is accelerated by high higher Nabundance, so zoned diamond (with a sawtooth variation in Nabundance) should trend towards an exaggerated sawtooth pat-

1 ica Act

tCdrpbnCbcvms

fiostdicdla

4

c“bldF1ftdflocdoacCAds2cscmdttdS

pdiRn

16 L. Barron et al. / Spectrochim

ern in terms of N aggregation. This was not found in zonedopeton/Bingara macrodiamond [10,14], implying that plasticeformation during growth was so severe that N aggregationeached some sort of coherent but extreme limit, apparently inde-endent of N abundance. This extreme limit was little influencedy subsequent events (storage, delivery). Determined from inter-al pressures on inclusions (Pr), 15 entrapment loci (6 from oneopeton stone MR2/8) intersect within the UHP macrodiamond sta-ility field [9, and unpublished results on diamond MR2/8]. For PThanges due to storage and delivery, this means that the relativeolume changes of the diamond/inclusion system must have accu-ulated elastically, indicating that plastic deformation was not a

ignificant process during either storage or delivery.The presence of singly substituted N is characteristic of diamond

rom exposed UHP terranes [6] because they were exhumed rapidlymmediately after termination of subduction. Hence the absencef singly substituted N in Copeton/Bingara diamond is indirectupporting evidence for volcanic delivery rather than exhuma-ion. Features directly indicating the high temperature of volcanicelivery include resorption of diamond with shiny surfaces featur-

ng microdiscs, etched percussion marks/fractures, and relativelyommon ruptured inclusion chambers [9,10,14]. Walcha districtiamonds are visually similar to those from Copeton/Bingara and

ack singly substituted N, but are distinct [16] with low N abundancend low levels of IaB aggregation.

.3. Raman spectra

Determinations of the 2nd order Raman spectra of Africanratonic diamond (Australian Museum collection) show that thecratonic” 2nd order Raman peak is reliably present as reportedy Menneken, but only if the diamond is non-fluorescent in UV

ight. The critical peak was completely suppressed in all Africaniamond tested that visibly fluoresced (for example top trace inig. 5a), equivalent to a Raman background higher than about500 counts per second. Thus Menneken’s test is only validor non-fluorescent diamond: 82% of the 27 African diamondsested are non-fluorescent, compared with 13% of 700 Copetoniamond tested. Henceforth the Menneken test is restricted to non-uorescent diamond: the S/N ratio ranges from 15× to 75× for 77%f 23 African cratonic diamonds (peak is relatively large, Fig. 5a). Inontrast (Fig. 5b), S/N ranges from 2× to 14× for 92% of 35 Copetoniamonds (peak is relatively small) and for each of a small numberf other diamonds tested (5 of 6 from Bingara, 3 of 3 from Walcha,nd 3 of 3 from Cempaka). The 92% level for Copeton diamondould actually represent 100% for UHP macrodiamond since 5–9% ofopeton (and Bingara) diamond was reported to be cratonic [10,12].lthough the 2nd order Raman spectra for this non-cratonic macro-iamond is considerably masked by noise, there are fine spectraltructures in common with cratonic diamond (definite peaks at450, 2640 and 2700 cm−1). For our use, a low S/N ratio for theritical peak is more reliable than the fine details of the 2nd orderpectra. The limited results for Bingara, Walcha and Cempaka indi-ate that non-cratonic macrodiamond is present in those regions,ostly likely of UHP origin like Copeton stones. Only two Argyle

iamonds out of more than 50 in the Australian Museum collec-ion are non-fluorescent, and one of these was tested by Raman:he 2nd order spectra is quite different from either cratonic or UHPiamond, with rounded rather than pointed peaks. However, the/N ratio is about 10×, suggesting a non-cratonic origin.

A fine crystal/grain size decreases the S/N ratio for all Ramaneaks, most visibly by increasing noise [29,30]. The influence of

iamond grain size on the Menneken test could be established

f a non-fluorescent cratonic diamond with a strong 2nd orderaman peak was crushed, and retested at various grain sizes. Men-eken’s compilation also demonstrated that the first order Raman

a Part A 80 (2011) 112– 118

peak position of non-cratonic microdiamond differed from cratonicmacrodiamond. Although a few Copeton macrodiamonds showedsimilar differences, it was a rare feature and was possibly due tointernal strain in the particular diamond.

4.4. X-ray: theoretical density of UHP macrodiamond

Values for the <2 2 0> and <3 1 1> peaks respectively are Cu K�-1 (75.2766◦, 91.4834◦) and Cu K�-2 (75.4972◦, 91.7791◦). Thesepeaks result in a calculated average value for the unit cell of theCopeton diamond of ao = 3.567 ± 0.001 A. Using [27]: D = (ZAC)/V,where A is the atomic weight of a formula unit (12.011), C is a con-stant (1.6601974), V is the volume per unit cell = a3

o, this equates toa calculated density D = 3.515 ± 0.001 g/cm3. The theoretical den-sity for Copeton diamond falls between the two density-groups[20] of diamond, so the theoretical density of Copeton diamondis normal. Six small diamonds were crushed to generate sufficientpowder for the X-ray treatment, yet the peaks were sharp, indicat-ing all shared similar degrees of crystal perfection, so this calculateddensity is taken as representative of Copeton UHP diamond. The oldreports of anomalous diamond density [1] must have some error inmeasurement.

4.5. Tectonic controls for UHP metamorphism: implications formacrodiamond

UHP metamorphism occurs when subducted rocks are draggedto depths great enough to reach the stability field of either coesiteor diamond (about 2.6 GPa at subduction temperatures). Mostknowledge about UHP metamorphism is derived from exposed(completely exhumed) UHP terranes. Considering that most sub-ducted rock is not exhumed, these exposed terranes are actuallyanomalous. They exhume to the surface as a terrane becausethey have a low density due to one or more of the followingfeatures: a high temperature/strong serpentinisation of entrainedmantle rocks/a low abundance of high density rocks (1–5% gar-net pyroxenite/eclogite). Some exposed classic UHP terranes carryUHP microdiamond (rapidly grown skeletal shapes, requiring ashort period within the diamond stability field, and invariably char-acterized by single substituted nitrogen demonstrating a shortperiod of mantle storage), whereas macrodiamond is absent toextremely rare [24]. However, graphitized macrodiamond is super-abundant (5–12%) in some exposed terranes (metamorphosedgarnet pyroxenite dyke in Ronda Peridotite [31]) and the forma-tion of the diamond correlates with subduction metamorphism[32], so this Ronda locality clearly falls within the concept of UHPmetamorphism. In this regard, it is vital to recognize that the con-trast between graphitization of macrodiamond versus survival ofmicrodiamond is not based on the exhumation PT path or rate.Microdiamond survival (in exposed UHP terranes) is due solely toprotection via the retained high pressure of supercritical fluid shar-ing the same inclusion chamber [33]. Thus the graphitization ofmacrodiamond in an exposed terrane should not exclude a terranefrom being classified as UHP.

The spectrum of UHP rocks can be broadened to include all sub-ducted rock that reaches about 2.6 GPa, regardless of the degreeof subsequent exhumation or re-equilibration. Within this widerscope, the following properties are presented as end membercontrols on UHP metamorphism: (1) ocean floor slab tempera-ture (cold/hot); (2) speed of subduction (slow/fast); (3) mantlegeotherm temperature (cold/hot); (4) timing of mantle serpentin-isation relative to subduction termination (early/late/never); (5)

strength of coupling between incoming continent and subductingslab (weak/strong); (6) density of continental margin (low/high).While some combinations have resulted in exposed exhumed UHPterranes lacking macrodiamond, other combinations could produce

L. Barron et al. / Spectrochimica Acta Part A 80 (2011) 112– 118 117

F itraryi eft panm aka (B

m“(rt

csGVVdat

tUCmf

5

udstdnUucmiopmmldPd

[

ig. 5. Raman 2nd order spectral peaks of uncut diamond, with the intensity in arbn cps. Non-fluorescent diamond has a Raman background less than 1500 cps. The l

acrodiamond from Copeton (130.22), Bingara (JB26), Walcha (D49910-12), Cemp

acrodiamond-bearing UHP terranes that stranded at depth. Thestranded” UHP model for the Copeton/Bingara macrodiamondspublished in [9,18] and summarized herein) has been included ineviews by leading researchers from the UHP field [34] and fromhe macrodiamond field [35,36].

When a continent breaks apart due to major rifting, the newontinental margin may contain up to 60% basaltic flows and intru-ions, defining a volcanic passive margin (VPM), for example thereenland coast facing Europe. If the oceanic plate attached to aPM is subducted, termination of subduction by collision with thePM would create a UHP terrane dominated by eclogite, with aensity high enough to encourage stranding in the upper mantlend lower crust. Such a UHP terrane would tend to remain withinhe diamond stability field, prolonging diamond growth.

A “lucky” combination of the above UHP end member con-rols represents the tectonic requirements of forming abundantHP macrodiamond (e.g. Copeton/Bingara). Volcanic delivery ofopeton/Bingara macrodiamond to the earth’s surface logicallyeans this combination would be mutually exclusive with that

orming microdiamond in exhumed UHP terranes.

. Conclusions

Macrodiamond from Copeton and Bingara shows distinctivenique features indicating a non-cratonic Phanerozoic UHP sub-uction origin. Differences in nitrogen abundance and aggregationtates between the two diamond parcels indicate multiple forma-ion regions, multiple local sources and the following compoundelivery mechanism: (1) on termination of subduction by conti-ental collisions, due to their high density, the diamondiferousHP terranes were stranded at depth, partially exhumed into thepper mantle and lower crust and (2) at some later time there wasapture and transport to the surface of UHP minerals (macrodia-ond, garnet, pyroxene) by deeply sourced post-arc alkali basaltic

ntrusions. The density of Copeton diamond is normal. Secondrder Raman peaks that are strong in cratonic diamond are sup-ressed in Copeton diamond. However, visible UV fluorescenceasks such peaks, so this non-cratonic diamond test (Menneken’s)ust be restricted to stones which are non-fluorescent. Similar but

imited Raman results suggest there is non-cratonic UHP macro-iamond at other places in NSW and in Kalimantan (Indonesia).ervasive deformation during growth of Copeton/Bingara macro-iamond causes: (1) X-ray crystal imperfection; (2) the relatively

[

[

units. The bold number adjacent to each trace represents the Raman backgroundel covers African diamonds (Australian Museum numbers). The right panel covers

.4, 6.8A), and Argyle (D49913G).

strong but coherent degree of N aggregation; (3) a UHP-reductionof the 2nd order Raman peak that otherwise is large for theoret-ical “perfect” and cratonic diamond; and (4) relative to cratonicdiamond, Copeton/Bingara diamonds have an incredible durabilityand refractory nature documented through their industrial use [37on p. 19].

Acknowledgements

The following sources are thanked for “non-destructive” accessto more than 3000 diamonds from known localities: G. White(Copeton, 96 and 130 series), and Copeton Bingara (P. Kennewell,Cluff Resources N.L., donated to the Australian Museum); Walcha(Australian Museum); Cempaka (Kalimantan, L. Spencer). D. Austenis thanked for cutting investigative windows on diamonds, andfor preparing crushed diamond for X-ray determinations. Thismanuscript is published with the permission of the CEO of Geo-science Australia.

References

[1] A.A. MacNevin, Diamonds in New South Wales. Geological Survey of New SouthWales, Mineral Resour. 42 (1977).

[2] E. Scheibner, Geology of New South Wales—Synthesis: v. 2. Geological Evo-lution: Geological Survey of New South Wales Memoir Geology, vol. 13 (2),Department of Mineral Resources, 1998.

[3] N.V. Sobolev, J.E. Glover, P.G. Harris (Eds.), Kimberlite Occurrence and Origin:A Basis for Conceptual Models in Exploration, vol. 8, University of WesternAustralia Publication, 1984, pp. 213–226.

[4] F.L. Sutherland, J.D. Hollis, L.R. Raynor, Mineral. Mag. 49 (1985) 748–751.[5] S.Y. O’Reilly, R.W. Johnson (Eds.), Intraplate Volcanism in Eastern Australia and

New Zealand, Cambridge University Press, Cambridge, UK, 1989, pp. 296–297.[6] W.R. Taylor, A.L. Jaques, M. Ridd, Am. Mineral. 75 (1990) 1290–1310.[7] F.L. Sutherland, P. Temby, L.R. Raynor, J.D. Hollis, in: H.O.A. Meyer, O.H. Leonar-

dos (Eds.), Proceedings of the 5th International Kimberlite Conference, Araxa,Brazil, 1991, vol. 2. Diamonds: Characterisation Genesis and Exploration,Companhia de Pesquisa de Recursos Minerais Brasilia, vol. 1B, CPRM SpecialPublication, 1994, pp. 170–184.

[8] L.M. Barron, S.R. Lishmund, G.M. Oakes, B.J. Barron, F.L. Sutherland, Aust. J. EarthSci. 43 (1996) 257–267.

[9] L.M. Barron, B.J. Barron, T.P. Mernagh, W.D. Birch, Ore Geol. Rev. 34 (2008)76–86.

10] H.O.A. Meyer, H.J. Milledge, F.L. Sutherland, P. Kennewell, Russ. Geol. Geophys.38/2 (1997) 305–331.

11] H.J. Milledge, F.L. Sutherland, P. Kennewell, 7th International Kimberlite Conf.Cape Town Ext. Abs., 1998, pp. 587–588.

12] R.M. Davies, The characteristics and origins of alluvial diamonds from EasternAustralia, Unpublished Ph.D. thesis, Macquarie University, Sydney, Australia,1998.

1 ica Act

[[[[[

[[

[[

[

[

[

[

[

[

[

[

[[[[[

18 L. Barron et al. / Spectrochim

13] W.L. Griffin, S.Y. O’Reilly, R.M. Davies, Rev. Econ. Geol. 11 (2000) 291–310.14] R.M. Davies, S.Y. O’Reilly, W.L. Griffin, Econ. Geol. 97 (2002) 109–123.15] R.M. Davies, W.L. Griffin, S.Y. O’Reilly, A.S. Andrew, Lithos 69 (2003) 51–66.16] F.L. Sutherland, L.M. Barron, Aust. J. Earth Sci. 50 (2003) 975–981.17] A.V. Korsakov, V.P. Zhukov, P. Vandenabeele, Anal. Bioanal. Chem. 397 (2010)

2739–2752.18] B.J. Barron, L.M. Barron, G. Duncan, Econ. Geol. 100 (2005) 1565–1582.19] M. Menneken, A.A. Nemchin, T. Geisler, R.T. Pidgeon, S.A. Wilde, Nature 448

(August) (2007) 917–922.20] Y.L. Orlov, The Mineralogy of the Diamond, Wiley Interscience, New York, 1977.21] A. Banas, T. Stachel, K. Muehlenbachs, T.E. McCandless, Lithos 93 (2007)

199–213.22] L.F. Dobrzhinetskaya, Z.H. Liu, P. Cartigny, J.F. Zhang, D. Tchkhetia, R.J. Hemley,

H.W. Green II, Earth Planet. Sci. Lett. 248 (2006) 340–349.

23] E. Thomassot, P. Cartigny, J.W. Harris, K.S. Vijoen, Earth Planet. Sci. Lett. 257

(2007) 362–371.24] K. DeCorte, A. Korsakov, W.R. Taylor, P. Cartigny, M. Ader, P. DePaepe, Isl. Arc 9

(2000) 428–438.25] H.M.J. Smith, Philos. Trans. R. Soc. A 241 (1948) 105–145.

[

[

[

a Part A 80 (2011) 112– 118

26] D. Strauch, W. Windl, H. Sterner, P. Pavone, K. Karch, Physica B 219–220 (1996)442–444.

27] J.J. Zussman, in: J. Zussman (Ed.), X-ray Diffraction. Physical Methods in Deter-minative Mineralogy, Academic Press, 1967.

28] A.S. Stepanov, V.S. Shatsky, D.A. Zedgenizov, N.V. Sobolev, Russ. Geol. Geophys.48 (2007) 758–769.

29] A.V. Spivak, Y.A. Litvin, A.V. Shushkanova, V.Y. Litvin, A.A. Shiryaev, Eur. J. Min-eral. 20 (2008) 341–348.

30] M. Perraki, D.C. Smith, E. Mposkos, Spectrochim. Acta A 68 (2007) 1077–1084.31] D.G. Pearson, G.M. Nowell, J. Petrol. 45 (2004) 439–455.32] L. Sanchez-Rodrigues, D. Gebauer, Tectonophysics 316 (2000) 19–44.33] L.M. Barron, Can. Mineral. 43 (2005) 203–224.34] J.G. Liou, W.G. Ernst, R.Y. Zhang, T. Tsujimori, B.M. Jahn, J. Asian Earth Sci. 35

(2009) 199–231.

35] J.G. Gurney, H.H. Helmstaedt, S.H. Richardson, S.B. Shirey, Econ. Geol. 105 (2010)

689–712.36] H.H. Helmstaedt, J.J. Gurney, S.H. Richardson, Can. Mineral. 48 (2010)

1385–1408.37] A. Joris, A Destiny in Diamonds, Boolarong Publications, 1986.