Fig. 3. U-Pb concordia diagram of zircon analyses from sample OM95-27, the felsic metavolcanic. Numbers refer to
analyses in Table 2. Error ellipses are 2s.
on several characteristics, as follows. Attempts were made to select zircons that were clear, crack-
free and lacking inclusions, cores and over- growths. However, the poor quality of samples
often made selection of cracked and cloudy grains necessary. Titanite fragments were divided based
on colour and selected for internal clarity and lack of cracking. Apatite grains are prismatic
whereas the feldspars are generally anhedral. Both were selected for clarity and lack of cracking.
Rutiles were chosen for crystal shape, colour and luster an indicator of freshness. Rutile aggre-
gates were washed in HF for 15 min in an ultra- sonic cleaner to remove surrounding silicate
minerals. Selected zircons were abraded Krogh, 1982, but many were left unabraded due to the
possibility of shattering because of internal cracks. No other minerals were abraded. Large
titanite, apatite and feldspar fractions were weighed, whereas weights for zircon, baddeleyite
and rutile fractions were estimated by eye and are likely accurate to about 9 50. Zircon, baddeley-
ite and rutile were digested in bombs, while titan- ite, apatite and feldspar were digested in Savillex
capsules. For zircon, U and Pb were separated using HCl with 50 ml anion exchange columns
following the method of Krogh 1973. Titanite, apatite, rutile and feldspar were passed through
500 ml anion exchange columns with HBr, follow- ing the method of Corfu 1988. Pb blanks are 1
pg for small column and 2 pg for large column chemistry. U blanks are taken to be 0.1 pg. U and
Pb were loaded onto Re filaments using H
3
PO
4
and silica gel. Isotopic analysis was carried out on a VG354 mass spectrometer in peak jumping
mode, with either a Faraday collector for large signal samples or Daly detector for small signal
samples. The mass discrimination correction for the Daly detector is 0.40 per AMU and the
thermal mass discrimination factor is 0.10 per AMU. Common lead in the apatite, titanite and
rutile samples made ages variably dependent on the isotopic composition defined for the initial
common Pb. Data from the feldspar from OM95- 2 were used as a common lead correction on all
regressions for these minerals. Data are plotted with two sigma error ellipses on U-Pb and Pb-Pb
diagrams Figs. 3 – 9. Regressions are calculated during pre-gold and gold-forming stages. Sul-
phides are mostly pyrite, but also include minor sphalerite, galena, and chalcopyrite. Pyrite miner-
alization occurred before, during and after gold deposition, as did silicification. Increased sulphi-
dation and silicification correlate with increased veining and gold mineralization.
A several hundred metres thick gabbro dyke, believed to be a member of the Avanavero Suite,
cuts the mineralized pluton at a depth of : 200 – 300 m Bertoni et al., 1991. Members of this
group of mafic rocks occur throughout the Guyana Shield in the form of sills, dykes and
other irregular bodies and have been dated by K-Ar, Rb-Sr and Ar-Ar methods. Sidder and
Mendoza 1995 estimated the age at between 1650 and 1850 Ma, based on compiled data from
several sources.
4. Analytical methods
Sample sizes ranged from 6 to 25 kg before crushing. Heavy mineral separates from the Wi-
lfley table were screened to − 70 mesh, and then separated using conventional magnetic and heavy
liquid techniques. Grains were hand-picked based
Fig. 4. Pb-Pb diagram showing feldspar analyses from OM95- 2, and hydrothermal titanite, apatite and rutile analyses from
OM95-2, OM95-26 and OM95-28, respectively. The error el- lipses are highlighted by circles. The line represents a regres-
sion of all analyses except 7 titanite and 29 feldspar. The Stacey and Kramers 1975 growth curve is also shown.
using the program of Davis 1982. Resulting age errors are quoted at 95 confidence levels.
5. Samples and results
Mineralized and
non-mineralized samples,
showing varying degrees of hydrothermal alter- ation, were chosen in order to obtain ages on
both igneous emplacement and hydrothermal crystallization Fig. 2. U-Pb analytical data for
samples with highly to moderately radiogenic Pb are presented in Table 2. Pb-Pb data for low
UPb titanite, apatite, rutile and feldspar fractions are presented in Table 3. Note that analyses 5 – 6
are given in both tables.
5
.
1
. OM
95
-
27
Meta6olcanic host rock Sample
OM95-27 is
an intermediate-felsic
metavolcanicsubvolcanic rock recovered as drill core from 130 m depth under the Wenot Lake
zone. It consists of plagioclase and minor quartz phenocrysts in a fine-grained quartz-feldspar ma-
trix. Secondary carbonate, chlorite and sericite are abundant, with minor pyrite and accessory zircon.
Zircons from this sample are elongate l:w 3:1, clear and colourless, with an average length
of : 50 mm. Two abraded grains were run as separate analyses, along with one unabraded piece
that contained longitudinal cracks. Data from the abraded fractions are concordant and, with the
slightly discordant datum from the unabraded fraction, define an age of 2120 9 2 Ma Fig. 3.
5
.
2
. OM
95
-
2
Marginal hornblende diorite Three samples of hornblende-rich marginal
phases were collected. OM95-2 is a porphyritic hornblende diorite taken from the north edge of
the pit. The hornblendes in this unit are stubby crystals showing a sub-parallel orientation. The
matrix consists of quartz and plagioclase with minor smaller amphiboles and homogeneously
distributed magnetite crystals. Strong hydrother- mal alteration produced carbonate, sericite and
epidote. Alteration of the hornblendes to chlorite and actinolite is pervasive. Accessory phases in-
Fig. 5. U-Pb concordia diagram showing igneous titanite and apatite analyses from OM95-2 and OM95-4, and hydrother-
mal titanite analyses from OM95-2. The line represents a regression of the igneous titanite and apatite results forced
through the origin of concordia.
C .
Norcross et
al .
Precambrian
Research
102 2000
69 –
86
Table 2 U-Pb data for zircon, baddeleyite, titanite and apatite
a
ThU Pb
com
pg
207
Pb
207
Pb
206
Pb Age
206
Pb
238
U
207
Pb
235
U Weight mg
Disc Number
U ppm Fraction
Ma
204
Pb OM
95
-
27
meta6olcanic 746.9
0.3894 9 8 7.068 9 20
0.002 2120.0 9 2.2
98 1 Ab z, flat
1 0.28
0.9 355.4
0.3891 9 12 2
7.060 9 28 1 Ab z, round
2119.6 9 4.2 0.1
0.001 84
0.29 0.8
669.7 0.3794 9 8
6.877 9 18 2117.8 9 2.2
2.5 5.0
1 z, frag, cracked 0.6
169 0.006
3 OM
95
-
2
hornblende diorite 165.5
0.3816 9 8 6.820 9 18
4 2092.9 9 1.8
Brown titanite 0.5
0.477 61
3.83 560
49.38 0.3580 9 13
6.096 9 29 2007.5 9 5.1
2.0 36
5 3.13
9.5 0.050
colourless titanite 1.69
266 32.70
0.3520 9 8 5.965 9 28
1998.6 9 6.3 3.2
6 0.290
colourless titanite 6.3
43.09 0.3652 9 8
6.365 9 24 2048.3 9 4.4
2.4 243
11 7
colourless titanite 0.226
2.28 OM
95
-
4
bladed hornblendite 79.01
0.3844 9 9 6.875 9 21
8 2094.2 9 2.7
Brown titanite -0.1
0.303 79
2.53 1101
417.7 0.3812 9 8
6.822 9 18 2095.4 9 1.6
0.8 517
9 1.48
397 0.181
Dark brown titanite 179
Colourless titanite 65.77
0.3764 9 9 6.730 9 23
2093.9 9 3.0 2.0
0.288 11
2.12 10
22.53 0.3697 9 9
6.592 9 56 2089 9 13
3.5 528
1.39 Apatite
0.225 6.0
11 OM
95
-
11
quartz diorite 1.64
12.1 67.75
0.3279 9 10 5.843 9 24
2087.6 9 4.2 14.3
0.001 12
228 3 Ab z, fuzzy core
180.9 0.3134 9 8
5.562 9 26 13
2080.7 9 7.0 2 z, cracked, fuzzy core
17.7 0.003
230 1.3
10.8 174.6
0.2900 9 6 5.148 9 24
2081.1 9 7.4 23.9
5.9 195
1.62 14
1 z, v. cracked 0.002
4.2 1 z, cracked
372.7 0.3456 9 10
6.163 9 24 2088.9 9 3.8
9.7 0.006
88 1.01
15 1.14
69.1 61.95
0.2879 9 8 5.001 9 60
2043 9 21 22.8
16 3 z, cracked
0.005 308
OM
95
-
28
quartz diorite 31.02
0.3095 9 14 5.512 9 44
2087 9 11 19.0
17 1 Ab z, clear
0.0002 261
1.43 8.8
47.92 0.3204 9 10
5.725 9 30 2092.5 9 5.8
16.4 9.2
221 1.28
18 1 Ab z, clear
0.0005 11.9
3 Ab z, fuzzy cores 100.3
0.2793 9 8 4.988 9 16
2091.3 9 2.8 27.1
0.002 214
1.42 19
1.24 9.3
214.7 0.2432 9 6
4.301 9 12 2074.0 9 1.8
35.9 20
4 z, cracked 0.003
304 OM
95
-
1
normal diorite 21
93.58 4 Ab z, tiny
0.3547 9 10 6.297 9 22
2081.1 9 3.6 6.9
0.001 267
1.78 10.2
1051 0.3072 9 6
5.291 9 14 2027.5 9 2.0
16.9 2.2
2 z, cracked 22
1.2 309
0.003 6.2
2 z, cracked 357.5
0.3189 9 8 5.541 9 20
2043.1 9 3.8 14.5
0.006 136
1.06 23
1.28 3.8
1224 0.3413 9 32
5.989 9 58 2060.6 9 4.4
9.4 24
7 z, cracked 0.01
163 OM
95
-
3
gabbro dyke 25
485.1 1 Baddeleyite
0.3166 9 14 4.766 9 23
1786.0 9 4.5 0.8
0.001 103
0.05 0.5
2099 0.3187 9 10
4.808 9 15 1790.0 9 2.5
0.4 0.7
1.29 26
1 Ab zr 0.0015
418 340.9
27 0.3403 9 14
1 Unab zr 5.240 9 22
1826.7 9 4.6 -3.9
0.004 763
1.37 21.5
a
Notes: Errors are two standard deviations. Pb blank–1 pg zircon, 2 pg titanite and apatite; U blank–0.1 pg. Ab, abraded, z, zircon. Pb
com
, Common Pb, including blank; calculated using blank isotopic composition. ThU calculated from radiogenic
208
Pb
206
Pb ratio and
207
Pb
206
Pb age assuming concordance. Disc. = per cent discordance for the given
207
Pb
206
Pb age. Decay constants are from Jaffey et al. 1971.
Fig. 6. U-Pb concordia diagram showing OM95-11 zircon analyses. The two sigma error ellipses are highlighted by
circles. The line represents a regression of all data except analysis 16. Only analysis 12 was from an abraded fraction.
Fig. 8. U-Pb concordia diagram showing zircon analyses from OM95-1 Diorite. The two sigma error ellipses are highlighted
by circles. The upper line 2096 9 11 Ma was regressed with- out analysis 21 the only analysis from an abraded fraction.
The lower line 2110 9 6 Ma is a regression through all four data.
Fig. 7. U-Pb concordia diagram showing zircon analyses from OM95-28 vein wall rock. The two sigma error ellipses are
highlighted by circles. The line represents a regression of all data. Analyses were from abraded fractions except for 20.
Fig. 9. U-Pb concordia plot showing analyses of zircon and baddeleyite from a late mafic dike. The line is a regression
through the three data.
clude pyrite and chalcopyrite, and titanite is found as small subhedral zoned crystals scattered
throughout the matrix.
Table 3 Pb isotopic analyses for secondary minerals from hydrothermally altered rocks at Omai
a
Fraction Number
Weight mg U ppm
PbCom pg Corr. Coef.
206
Pb
204
Pb
207
Pb
204
Pb
208
Pb
204
Pb OM
95
-
2
Hornblende Diorite Colourless
5 0.050
9.5 36
0.9989 306.0 9 16.4
295.4 9 14.8 51.2 9 2.0
titanite Colourless
0.290 6.3
266 6
0.9491 157.74 9 1.05
104.01 9 0.76 32.81 9 0.15
titanite 7
Colourless 0.226
11.0 243
0.9744 236.82 9 1.78
43.30 9 0.25 179.5 9 1.3
titanite 0.553
0.16 3180
0.9781 28
15.682 Feldspar
15.356 35.55 9 0.12
9 0.036
9 0.024
0.29 31
7415 0.8606
29 17.595
Feldspar 15.491
40.30 9 0.16 9
0.040 9
0.034 OM
95
-
26
Quartz Diorite 0.175
30 1.6
3224 0.9838
17.010 15.518
Apatite 36.44 9 0.11
9 0.036
9 0.027
OM
95
-
28
Quartz Diorite 0.048
4.8 31
766 0.9338
21.430 54.30 9 0.19
16.042 Rutile, dark
9 0.038
9 0.039
0.402 6.6
5042 0.9693
Rutile, dark 25.981
16.635 32
82.70 9 0.26 9
0.039 9
0.043 16.694
0.683 24.0
30426 Rutile, yel-
0.9971 26.465
33 64.45 9 0.23
9 0.041
low 9
0.038
a
Footnotes to Table 2 apply.
Two fractions of plagioclase feldspar were analysed from this sample, to be used for com-
mon lead corrections. Fig. 4 insert shows the Pb-Pb plot of the two points, with the Stacey and
Kramers 1975 curve for reference. Analysis 28 Table 3 contains the most primitive Pb and plots
just above the curve, whereas analysis 29 plots below it but shows much more evolved
206
Pb
204
Pb. The more radiogenic Pb data from other minerals plot colinearly with datum 28, but do
not regress successfully with the more evolved feldspar datum 29. The feldspar fractions were
picked to be as fresh as possible but slight turbid- ity was evident in many of the grains, indicating
the presence of some alteration. The high U con- tent of analysis 29 Table 3 suggests that it was
from an impure feldspar fraction that may not have
preserved a
primary Pb
isotopic composition.
Analysis of one fraction of brown titanite gave a data point only 0.5 discordant with a
207
Pb
206
Pb age of 2093 9 2 Ma when corrected for the initial composition of the feldspar analysis 4, Fig.
5. The brown titanite is relatively high in U and its Pb is fairly radiogenic. Three fractions of
colourless titanite fragments were also analysed from this sample. These contained much lower U
concentrations and had less radiogenic Pb. Using the feldspar datum to correct for common Pb, the
data points are 2 – 3 discordant and show vari- able
207
Pb
206
Pb ages of 2048 9 4, 2008 9 5 and 1999 9 6 Ma Table 2. The two younger data
regress within error of the feldspar datum on a Pb-Pb diagram analyses 5 and 6, Fig. 4, and also
regress within error of data from hydrothermal rutile and apatite from other samples see below.
The 2048 Ma titanite datum lies distinctly off this line and is probably the result of inadvertently
mixing the younger hydrothermal or metamorphic titanite with older igneous titanite, which some-
times forms cores of crystals. Because of the relative U concentrations, any accidental mixing
of phases would bias the result disproportionately toward the age of the older component.
5
.
3
. OM
95
-
26
Altered hornblendite OM95-26 is a hornblendite with strong hy-
drothermal alteration chosen from the southeast hornblende-rich margin of the pluton. It contains
bladed amphibole crystals that are pseudo- morphed by fine-grained carbonate, sericite, chlo-
rite and some minor plagioclase and opaque minerals including pyrite with microscopic gold
inclusions, magnetite and chalcopyrite. The ma- trix consists mostly of quartz and plagioclase,
usually in a granophyric intergrowth. The quartz is filled with H
2
OCO
2
fluid inclusions and the plagioclase is strongly altered to fine-grained
sericite and calcite. Apatite crystals were found in the matrix, as were fine rutile needles and
pleochroic haloes indicating possible zircons.
Apatite was the only potentially datable min- eral that could be recovered from this sample. A
fraction of well-formed, clear apatites : 250 mm long was analysed. The resulting datum analysis
30, Fig. 4 insert is non-radiogenic and plots close to the feldspar data from OM95-2.
5
.
4
. OM
95
-
4
Marginal hornblendite OM95-4 is a hornblendite taken from the west
side of the pit, where field relations show that it crystallized in situ. Large, zoned and twinned
hornblende blades that are partly altered to chlor- ite and actinolite make up 75 of the sample. The
matrix is quartz and albite, occasionally with granophyric texture, and the sample shows mod-
erate hydrothermal alteration to carbonate, epi- dote and sericite. Opaque minerals include pyrite,
magnetite, chalcopyrite and bornite. Zircon was not seen in thin section. Accessory minerals in-
clude large, zoned euhedral titanites, associated with magmatic hornblende that grew in situ in-
ward from the intrusive contact, and large euhe- dral apatite. Both appear to be of the same
generation as the hornblende.
Fragments of titanite up to 150 mm in size were recovered. Dark red-brown fragments were from
the cores of crystals, colourless fragments were from the edges, and mottled, medium-brown frag-
ments were presumably from areas between the centres and the edges. One multi-grain fraction of
each colour was analysed Table 2. Uranium contents increase with depth of colour. One data
point is within error of concordia while the others are only slightly discordant, but the
207
Pb
206
Pb ages are essentially the same and average to
2094 9 1 Ma Fig. 5: 57 probability of fit. One fraction of apatite crystals was also
analysed from this sample. Grains chosen for analysis were generally colourless, well-formed
prisms : 300 mm long, with slightly pitted sur- faces. Correcting for initial common Pb using the
feldspar datum from OM95-2 gives a
207
Pb
206
Pb age of 2089 9 13 Ma. The data point on a concor-
dia diagram is several percent discordant Fig. 5.
5
.
5
. OM
95
-
11
Southeast quartz diorite Sample OM95-11 is a quartz diorite with strong
hydrothermal alteration from the hanging wall of a major quartz vein system in the south end of the
pit. The matrix consists of quartz and plagioclase with chlorite, sericite, carbonate and rutile as
alteration minerals. Opaque minerals include pyrite and minor magnetite, with the former often
containing blebs of gold 30 mm in size. Zircons were found in thin section only as very small
grains 30 mm long in chlorite, and often only by their pleochroic haloes so it is unclear if the
grains seen in thin section are the same generation as those recovered by mineral separation.
Zircons recovered from this sample are smaller than those from the metavolcanic unit, colourless,
and often cracked with cloudy interiors. Five fractions were dated, but only one was abraded
due to the poor quality of the grains. The abra- sion did not improve concordance. One highly
discordant datum that is not colinear with the rest was not used in the regression Fig. 6. The lead
loss line gives an age of 2094 9 6 Ma, with a 50 probability of fit and a lower intercept close to
zero.
5
.
6
. OM
95
-
28
Central quartz diorite OM95-28 is vein wall-rock collected in the mid-
dle of the pit, in an area of high grade mineraliza- tion. It consists of plagioclase, carbonate and
quartz, with minor apatite, chlorite and pyrite.
The vein material directly adjacent to the sample contains visible gold. Zircons about 50 mm in size
were found in thin section. OM95-28 also con- tained rutile crystals large enough to be recovered.
In thin section rutile is usually found as clusters of striated, twinned crystals, dark brown to reddish
brown in colour, inside or bordering the matrix quartz and plagioclase.
Four fractions of colourless, slightly cracked zircons, three of which were abraded, were
analysed. All data are highly discordant, with the unabraded fraction giving the most discordant
datum, which is not quite collinear with the other three. The three most concordant data were used
to plot a Pb-loss line. The upper intercept with concordia is 2092 9 11 Ma, with a 36 probabil-
ity of fit and a near-zero lower intercept age Fig. 7.
Rutile was recovered from this sample as small 0.2 – 0.25 mm bundles consisting of brown elon-
gate crystals, some of which show elbow twinning. Rutile has previously been used to provide mini-
mum age constraints on mineralizing or signifi- cant hydrothermal events e.g. Davis et al.,
1994b. Three fractions were separated for analy- sis. The freshest looking consisted of ten clusters
of euhedral dark brown crystals with shiny stri- ated faces. A second fraction consisted of a large
number of similar clusters of brown crystals that were less shiny and euhedral than the first frac-
tion. The third fraction consisted of anhedral yellow-brown rutile. Pb from the rutile is only
moderately radiogenic. On a U-Pb concordia dia- gram, the data points are very imprecise because
of the small proportion of radiogenic Pb. On a Pb-Pb diagram the three analyses 31 – 33, Table
3 define an age of 2020 9 56 Ma when regressed together with the most primitive feldspar datum
Fig. 4, insert.
5
.
7
. OM
95
-
1
North diorite OM95-1, is an altered diorite from the north
end of the pit. It consists mostly of sericitized plagioclase, quartz, chlorite and epidote. Opaque
minerals include magnetite and pyrite, with no observable gold. The zircons found in thin section
are similar to those in OM95-11 i.e. they are extremely small and were seen in thin section only
when they formed a pleochroic halo in chlorite. Zircons recovered from this sample are mostly
small, cloudy and cracked. Discordant data from the three unabraded fractions define a Pb-loss line
with an upper intercept of 2096 + − 11 Ma Fig. 8: 13 probability of fit. The abraded fraction
gives the most concordant datum but is not colin- ear with the three other data points. When re-
gressed with the two most discordant points, this datum defined an age of 2110 9 6 Ma Figs. 8 and
73 probability of fit. The lower intercepts for the regressions are between 500 and 600 Ma.
Abrasion of cracked and altered zircons can produce scattered or even reversely discordant
data, possibly due to within-grain Pb and U mo- bility in altered domains followed by U-Pb frac-
tionation when part of the grain is removed by abrasion Davis et al., 1982. Therefore, the inter-
cept age determined from the three unabraded fractions is considered to be the most reliable in
spite of its lower probability of fit.
5
.
8
. OM
95
-
3
Gabbro dyke Sample OM95-3 is a drill core sample of the
gabbroic dyke that cuts through the Omai pluton at depth. The sample was obtained from beneath
the main pit. In thin section no dateable minerals were seen, although a few pleochroic haloes were
observed, indicating the possible presence of zircon.
Mineral separation yielded only a few zircon and baddeleyite grains. The zircon mostly consists
of cracked and altered elongate grains, whereas the baddeleyite grains are very tiny. Two zircons
were analysed from this sample, a crack-free grain that was abraded prior to analysis, and a second
elongate, prismatic, cracked crystal that was not abraded. One flat, fresh-looking grain of badde-
leyite was also analysed. The data from the un- cracked abraded zircon and the baddeleyite
Table 2 are nearly concordant, whereas the da- tum from a cracked unabraded zircon analysis
27 is reversely discordant Fig. 9. The regression of the three data points defines an age of 1794 9 4
Ma, with a 94 probability of fit.
Reversely discordant data have only rarely been observed from zircon. Lack of equilibration be-
tween spike and sample can produce spurious reversely discordant data, but this is unlikely to be
the case with analysis 27 because its
207
Pb
206
Pb age is older than those of the near concordant
data and the three data define a statistically con- sistent regression with a fairly old lower concordia
intercept age. The reversely discordant grain showed a relatively high U concentration, and
lower concordia intercepts near 1000 Ma are typi- cal for high U zircons from mafic rocks, probably
because they begin to accumulate radiation dam- age-induced Pb loss earlier than low U zircons. It
appears that this may be a real case where U loss from zircon exceeded Pb loss. Unfortunately, the
few remaining zircon grains from this sample are quite altered and, since they probably underwent
pronounced secondary Pb loss, they are likely to give complex discordia that would not clarify the
cause of the reverse discordance.
6. Discussion