SL dyke,..what is it?
http://www.cg.nrcan.gc.ca/slave-kaapvaal-workshop/abstracts/mclean.pdf
Abnormally thick Cambrian lithosphere of the Southeast Slave Craton: evidence
from crystalline inclusions in diamonds and pyrope compositions in Snap Lake
kimberlites.
N.P. Pokhilenko 1,2 , J.A. McDonald 1 , A.E. Hall 1 , N.V. Sobolev 2
1. Diamondex Resources Ltd. Vancouver B.C. Canada.
2. Institute of Mineralogy and Petrography, Novosibirsk, Russia.
Previous petrological and mineralogical studies of the Slave Craton kimberlites produced
a number of results suggesting a significant increase in lithosphere thickness towards the south of
the craton (Pokhilenko et al., 1997,1998,2000,2001; Griffin et al., 1999; Grutter et al., 1999;
Kopylova et al., 1999, 2001; Kopylova and Russell, 2000). In support of this inference, pyropes
from the Snap Lake (SL) kimberlite dyke system have an unusually wide range of Cr2O3 content
(up to 17wt.%). These high Cr2O3 concentrations in pyrope are consistent with the existence of a
very thick lithosphere beneath this region at the time of kimberlite emplacement (Pokhilenko et
al., 1998; 2000; McLean et al., this volume). Evaluation of pyrope compositions of the Slave
Craton kimberlites demonstrated a progressive increase of maximum Cr2O3 content in pyropes
from north to south in the craton (Grutter et al., 1999). Petrological studies of upper mantle
xenoliths permit the conclusion that lithosphere thickness was 160-190 km in the northern part of
the Slave Craton (Kopylova et al., 1999), ~ 200 km for the central Slave (Pearson et al., 1999) in
Jurassic-Tertiary time, and a minimum of 230 km in the southern (Kennady Lake area) part of the
Slave Craton (Kopylova et al., 2001). Data obtained from a mineralogical study of the Snap Lake
(SL) kimberlite indicate that the lithosphere thickness beneath this area was as much as 300 km
(Pokhilenko et al., 1998,2000; McLean et al., this volume) at the time of kimberlite emplacement
which is shown to be Cambrian time (Agashev et al., this volume) which is characteristic for the
South Slave Craton (Heaman et al.,1997).
We obtained preliminary results from crystalline inclusions in diamonds recovered from
the Snap Lake kimberlite. 104 of 109 diamond crystals studied contain mineral inclusions with
U-type parageneses and only 5 diamonds contain inclusions of E type parageneses (see
table)
Association N of diamond crystals
(N of diamonds with
sulphide inclusions)
Ol 70 (9)
En 10 (0)
CrPy 4 (1)
Chr 3 (2)
Ol + En 7 (0)
CrPy + Ol 4 (0)
Cpx + En 1 (0)
CrPy + Ol + En 2 (0)
Sulphide 2
Total U-type 104 (14)
E-type Cpx 3 (0)
Ga + Cpx 2 (1)
Total E-type 5 (1)
Total U + E 109 (15)
/ E-type, % 95.4
The most important results were
obtained from the garnet inclusion study.
Twelve diamonds contain garnet
inclusions, four of which contain a
significant majorite component. Both E-type
garnet inclusions contain this
component (5.0 and 6.1mol%). Two Cr-rich
pyropes (11.8 and 12.8 wt.% Cr2O3)
are subcalcic (4.68 and 5.11 wt.% CaO)
and have high Mg (Mg# 85.2 and 83.2).
These Cr-rich pyrope inclusions, both of
harzburgite paragenesis, contain 11.6
and 16.8 mol.% majorite component
respectively.
These two Cr-rich subcalcic pyropes are definitely related to depleted ultramafic rocks of
the lithospheric mantle. From experiments modeling natural ultramafic systems,
pressures of at least 110 kbar are required to achieve ~ 16-17 mol % dissolution of
majorite component into magnesian garnets (Irifune, 1987; Irifune, Ringwood, 1987).
This implies that at the time of SL kimberlite emplacement, the lithosphere thickness
beneath the area was over 300 km. Other indications that some SL diamonds formed
under very high pressures include: a) Cpx inclusions that contain high maximum K2O
contents in both E-type (up to 1.37 wt.%) and U-type (0.71wt.%) paragenesis; b) Two E-type
garnet inclusions in diamonds that contain high Na2O (0.33 and 0.38 wt.%) Of some
significance is the observation that in one of these garnets, the Na content (0.046 for 12
atoms of O) is significantly higher than Ti (0.024 for 12 atoms of O). Furthermore, an
absence of P (<0.001 for 12 atoms of O) suggests that the reaction R 2+ Al ↔NaSi, may
occur accompanied by partitioning of Si into octahedrally coordinated sites.
Associations and compositions of U-type inclusions within SL diamonds together
with data on pyrope composition contained within SL kimberlites suggest a significantly
lower degree of both depletion and differentiation of ultramafic material comprising
lithospheric mantle beneath this region. Relatively low depletion is indicated by
relatively low average Mg# for olivine, enstatite and pyrope inclusions in SL diamonds as
well as by a significantly higher average CaO content for Cr-pyrope inclusions (x=
4.6wt.% CaO) compared to CaO contents of Cr-pyrope inclusions in Siberian and South
African diamonds. Relatively minimal differentiation in the lithospheric mantle is
indicated by: a) SL U-type pyropes show a higher proportion of lherzolitic relative to
both harzburgite and wehrlite compositions than pyropes contained in kimberlites of
similar diamond grade from Siberia and South Africa; b) Very uniform compositions of
olivine and enstatite inclusions as well as a comparative abundance of enstatite inclusions
in SL diamonds that are associated with a range of other silicate minerals. The studied
olivine and enstatite inclusions in SL diamonds are relatively enriched in Fe (xMg# =
92.0%). Relatively low equilibrium temperatures of diamondiferous lithospheric mantle
peridotites beneath the SL area are indicated by low contents of CaO in olivine (0.01-
0.05; x=0.021wt.%) and in enstatite (0.16-0.42, x=0.31wt.%) together with data pointing
to very high maximum pressures of SL diamond formation.
These results allow us to propose that an abnormally thick lithosphere existed
beneath the Snap Lake area in Cambrian time. Depleted ultramafic rocks strongly
predominated in a thick interval from the Moho surface to at least 300 km at which
depths diamondiferous harzburgites and depleted lherzolites contained Cr-pyropes with
significant amounts of majorite component. This suggests that the total thickness of
diamondiferous peridotites “sampled” by the SL kimberlite is not less than 170km.
Evidence suggests that this lithospheric mantle was significantly less depleted, less
differentiated, and cooler (heat flow as low as 31-33 mWm -2 ) than lithospheric mantle
underlying either the Siberian or Kaapvaal cratons.
The small proportion of E-type mineral inclusions in SL diamonds (4.6%)
together with data indicating the absence of eclogites among the studied upper mantle
xenoliths from the 5034 Pipe, Kennady Lake cluster (Kopylova et al., 2001) suggests that
the subduction model (Griffin et al., 1999) is problematic at least for this part of the Slave
Craton. However, a significant number of yellow and grey cubes less than 2mm in size
in the SL diamond population may be E-type. This problem requires a more detailed
study. Finally, the existence of an unusually thick, relatively cool, diamondiferous
peridotitic zone in the lithospheric mantle underlying the Snap Lake area in Cambrian
time together with geochemical and isotopic data suggesting that the SL kimberlite melt
resulted from a small degree of partial melting of slightly carbonated depleted lherzolites
within the lithospheric mantle (Agashev et al., this volume) makes this part of the Slave
Craton extremely promising for diamond exploration. This reinforces earlier conclusions
of other authors (Pokhilenko et al. 1998; 2000; 2001; Kopylova et al.,2001). |
