HPHT treatment of CO2 containing and CO2-related brown diamonds

Research areas:
  • Thomas Hainschwang
  • Franck Notari
  • Emmanuel Fritsch
  • Laurent Massi
  • Benjamin Rondeau
  • Christopher M. Breeding
  • Heiner Vollstaedt
Ten type I diamonds containing CO2 and three diamonds related to these
(referred to as ``Pseudo CO2{''} diamonds) were treated by the HPHT
process to observe changes in colour, colour distribution, inclusions,
luminescence and spectral features in the visible to the mid-infrared
regions. All samples were of predominantly brown colour before the
treatment and had rather inhomogeneous colour distribution not related
to strain like in classic type la brown diamonds. Upon HPHT treatment
most of them changed to a more yellow hue, with a greenish modifying
colour, but the dominant colour was still brown. The modification of
colour was found to be much more discreet than it is the case for
deformation-related type la brown diamond and the colour was not at all
related to known centers such as H3. The colour modifications were
caused by the destruction of broad absorption bands in the Vis-NIR
spectrum by the HPHT process and generally increased transmission from
400 to 700 turn, likely related to a slight increase in single nitrogen,
but no typical HPHT annealing related colour centers such as H3 and H2
were produced in the spectra of the samples. This could be confirmed by
the luminescence when excited by shortwave and longwave UV light, under
which all samples showed a yellow emission that was practically not
modified by the treatment.
The infrared spectra of the 10 CO2 diamonds, defined by the presence of
at least the CO2 stretching (V3) and bending (V2) absorptions, were
unusually modified by the HPHT treatment: the treatment caused the CO2
bands to increase in intensity in diamonds with low to medium CO2
content, and the apparently diminished the intensity of the
nitrogen-related one phonon absorptions (when present). In the spectra
of the three ``Pseudo CO2{''} samples, which lack by definition the CO2
peaks, the CO2 peaks were created and the one phonon absorption reduced
by the treatment. The photoluminescence spectra exhibiting several
unpublished emission peaks remained practically unchanged by HPHT
processing: none of the emissions were annealed out and no new centers
The data obtained by HPHT treatment leads us to propose, that the theory
of solid CO2 inclusions being responsible for the C02 absorptions in
diamond may be erroneous, and that the CO2 may be present structurally.
Thus it is proposed that structurally bonded oxygen may play an
important role in such diamonds.
The present study indicates that treated CO2 and ``Pseudo CO2{''}
diamonds can in most cases not be identified by the methods available in
well-equipped gemmological laboratories, since no detectable defect
centers appear to be induced by the treatment. (C) 2008 Elsevier B.V.
All rights reserved.