Anal Bioanal Chem (2007) 387: 791–801
DOI 10.1007/s00216-006-0684-y
ORIGINAL PAPER
Francesca Casadio . Janet G. Douglas .
Katherine T. Faber
Noninvasive methods for the investigation of ancient Chinese
jades: an integrated analytical approach
Received: 30 May 2006 / Revised: 5 July 2006 / Accepted: 14 July 2006 / Published online: 16 August 2006
# Springer-Verlag 2006
Abstract This paper reports on an integrated analytical
approach for the noninvasive characterization of Chinese
nephrite samples, encompassing both geological reference
specimens and museum objects. Natural variations induced
by cationic substitutions, as well as human-induced
alterations such as heating, which both affect color, are
the focus of this contribution. Totally noninvasive methods
of analysis were used, including X-ray fluorescence
spectroscopy, Raman microspectroscopy, visible reflec-
tance spectroscopy and X-ray diffraction; moreover, the
feasibility of using a portable Raman spectrometer for the
in-field identification of jades has been demonstrated. Fe/
Fe+Mg (% p.f.u.) ratios of the jades have been calculated
based on hydroxyl stretching Raman bands, which will
provide an important addition to similar data that are being
collected at major museums in the Western and Eastern
hemispheres.
Keywords Jade . Nephrite . XRD . Raman spectroscopy .
Visible spectrophotometry
Introduction
In the early history of China, jade was highly regarded and
used extensively for ritual objects, decoration, utilitarian
objects such as tools, and burial goods. Quality and color
(termed Yu De and Yu Fu in ancient Chinese) were the two
fundamental parameters used to judge the “virtue” of jade
[1]. Early jades from China are typically composed of
nephrite, a massive, fine-grained variety of the tremolite–
actinolite series of amphiboles (Ca2Mg5Si8O22(OH)2 and
Ca2MgFe4Si8O22(OH)2, respectively) that consists of
interlocking fibrous crystals, although other stone materials
were used as well.
Research using scientific methods performed over the
last decade has focused on the mineralogical identification
of these materials, as well as the geological source of jade,
early jade working methods, the detection of heating in
jade, burial alteration, and surface accretions, using a wide
range of analytical methods [2]. In particular, noninvasive
techniques are finding support in studies of significant
collections of Chinese jades dating from the Neolithic
period (5000 to 1700 BCE) to the Han dynasty (206 BCE
to 220 CE) at major art museums. These studies help the art
historian to develop the cultural and archaeological
contexts of jade objects. Studies of unearthed jades, in
particular, are invaluable for determining the cultural
context of jade of uncertain origin, which is the situation
with most jades in Western collections [3–7].
Spectroscopic methods of analysis have been increas-
ingly applied to the study of jades, most commonly for
basic mineralogical identification [8, 9]. Other research has
aimed at answering questions beyond the mineral content.
New Zealand jade composed of nephrite has been
investigated by a combination of infrared, optical absorp-
tion spectroscopy, Mössbauer spectroscopy and other
chemical and petrologic methods in order to better
understand the color of nephrite from that region [10]. Of
particular interest is the use of noninvasive Raman
microscopy (RM) in the study of Mesoamerican jadeite
pebbles from Guatemala, where the mineral phases present
in the jade were used to classify the rock type [11]. Data
F. Casadio (*)
Department of Conservation Science,
The Art Institute of Chicago,
111 S. Michigan Ave,
Chicago, IL 60603, USA
e-mail: fcasadio@artic.edu
J. G. Douglas
Department of Conservation and Scientific Research,
Freer Gallery of Art and Arthur M. Sackler Gallery,
Smithsonian Institution,
Washington, DC 20560, USA
K. T. Faber
Department of Materials Science and Engineering,
Northwestern University,
Evanston, IL 60208, USA
from this study will continue to be useful for geologic
sourcing studies when compared to compositional data of
Pre-Columbian jade artifacts [12]. The application of RM
to jadeite is well demonstrated, and could become a routine
approach in archaeometry for identification and prove-
nance studies, especially as inexpensive portable Raman
microprobes are developed with improved spectral
resolution (up to 8 cm−1).
X-ray fluorescence spectroscopy (XRF) has been
applied to the study of Chinese jades composed of nephrite
in order to determine the minor elemental compositions of
these objects, and hence, more about their geologic source
in China [13, 14]. A combined, noninvasive, external beam
particle-induced X-ray emission (PIXE) and RM study has
been applied to six ancient Chinese jades [15]. This work,
performed at the Guimet Asian Museum in Paris, was
aimed at the determination of cation composition and
distribution in nephrite, which served as a way to
understand the color in jade and the provenance of ancient
Chinese jades. Recently, approximately twenty ancient
Chinese nephrite objects from the collection of the British
Museum in London were studied with RM and scanning
electron microscopy with elemental microanalysis (SEM/
EDX) to address similar questions [16].
The current study builds on this research, and focuses
specifically on the use of noninvasive methods that are
ideally suited to the study of ancient Chinese jades, because
the analysis can be done without risking damage to
precious artifacts, and several areas on the surface of an
object can be probed, thus allowing comprehensive
mapping of the phases present. In addition to determina-
tions of mineral content, the methods are useful for
studying the chemical causes of color in nephrite,
determining heating in jade, and further surface character-
ization. Additionally, the successful demonstration of the
applicability of a portable micro-Raman spectrophotometer
to the precise characterization of these materials will be
valuable for the study and classification of excavated jades
at archaeological sites in China. Together these techniques
provide us with a powerful tool for characterizing jade
materials used by people in ancient China, and for building
a database that will allow the study of unprovenanced
Chinese jades.
Experimental methods
The objects examined are described in Tables 1 and 2 and
Figs. 1 and 2. They include modern nephrite reference
specimens of known geological origin from China and
neighboring regions (a possible source of jade in ancient
times), as well as selected museum objects dated from the
Neolithic (third millennium BCE) to the Shang dynasty
(second millennium BCE). All of the objects were selected
on the basis of color, ranging from translucent white to
almost black.
To study the effect of heat treatment, a low-iron
tremolitic nephrite pebble from Hetian (Khotan) in
Xinjiang Province was cut with a diamond saw into
different slices, each polished on one side and individually
heated in a muffle furnace in oxidizing atmosphere for
24 hours at temperatures ranging from 500 °C to 1100 °C in
increments of 100 °C, as already previously described in
detail [17]. One unheated sample was kept as a control.
Table 1 Modern reference geological samples of nephrite from China and neighboring regions, arranged from lightest to darkest perceived
color
Sample Provenance Description Color
(CIELab
coordinates)
λ of major
absorption
vis
spectrum
(nm)
λmax
R%
(nm)
Apparent composition
(Raman)
(Fe/Fe +
Mg)(% p.f.u.)
Elemental composition detected by XRF (areas of normalized peaks, counts × keV)
Fe
(Kβ1,2)
Mg
(Kα1,2)
Zn
(Kα1,2)
Mn
(Kα1,2)
Ni
(Kα1,2)
Cr
(Kα1,2)
Sr
(Kα1,2)
Ti
(Kα1,2)
K
(Kα1,2)
5 Chuncheon,
South Korea
light grayish
white
L* 42.75 622 459 1.09 1.39×104 4.75×103 3.86×103 2.24×104 – 1.28×103 7.09×103 – –
a* −1.86
b* −3.34
7 Xiaomeiling,
Kiyang county,
Jiangsu province, China
light greenish
white
L* 73.72 640 532 1.64 1.16×104 4.69×103 6.11×103 5.26×103 – 1.36×103 7.85×104 1.62×103 5.04×103
a* −4.77
b* 4.39
2 Liaoming Province, Xiuyan
county, China
light greenish
yellow
L* 57.12 624 500 0.61 2.14×104 6.38×103 1.36×104 7.92×103 – 2.86×103 – 2.30×103 2.55×103
a* −3.62
b* 0.85
9 Tao river Valley,
Lintao, Gansu
province, China
light green L* 55.60 621 521 1.37 2.62×104 5.62×103 9.87×103 2.40×103 – 1.28×103 – – 4.28×103
a* −10.27
b* 9.10
11 Tao river Valley,
Lintao, Gansu
province, China
greenish yellow L* 29.55 629 499 0.64 1.59×104 5.05×103 5.17×103 6.46×103 – 1.32×103 – 2.90×103 6.80×102
a* −4.66
b* 2.26
3 Hualien, near
Fentien, Taiwan
olive green L* 29.40 630 528 13.23 2.02×105 3.41×103 6.79×103 2.13×104 4.61×104 1.59×104 – – –
a* −2.99
b* −0.39
10 Tao river Valley,
Lintao, Gansu
province, China
brownish green L* 15.29 620 513 3.06 7.37×104 4.79×103 6.51×103 1.23×104 – 1.28×103 – – 4.62×102
a* −5.06
b* 4.29
792
A Rontec (Bruker, Berlin, Germany) ArtTAX noninva-
sive portable microfocus XRF system with an Mo excita-
tion tube was employed, with a voltage of 50 kV, a current
of 800 μA and He flushing to enhance the detection of low-
Z elements. A live time of 200 seconds was used for all
geological nephrite specimens analyzed, so that after curve
fitting and normalization for the Ca peak, areas of all peaks
could be calculated and the relative abundance of each
element could be comparatively evaluated in a semiquan-
titative way.
Visible spectrophotometry was performed with a
SpectroEye spectrophotometer (GretagMacbeth, New
Windsor, NY, USA) over the range 380 nm to 730 nm
with 10 nm resolution; 45°/0° geometry and a gas-filled
tungsten type A illumination source were used. Data are
presented for illuminant C and a 2° observer.
Most Raman microscopy (RM) data were acquired using
a Jobin Yvon Horiba (Edison, NJ, USA) Labram 300
confocal Raman microscope , equipped with three excita-
tion laser lines (λ0=532 nm; 632.8 nm and 785.7 nm).
Power at the surface of each jade was kept low in order to
avoid thermal damage (measured at 0.94 mWand 3 mW for
532 nm excitation at two different neutral density filters
settings, with an 100× microscope objective). A battery-
powered Raman spectrometer (model Inspector Raman,
with a 70 mW diode laser excitation of λ0=785 nm, and
10 cm−1 resolution, manufactured by DeltaNu (Laramie,
WY, USA, see http://www.deltanu.com) was also used.
X-ray diffraction (XRD) of the nephrite specimens and
Chinese jades was carried out on a Scintag Diffractometer
(Scintag Inc., XDS 2000, Cupertino, CA, USA) with Cu
Kα radiation at 40 kV and 20 mA. Each spectrum was
collected from 8° to 70° 2θwith a step size of 0.1° 2θ and a
count time of 3 s. Surfaces used for diffraction were either
polished or naturally flat. Silicon powder was dispersed on
each surface as an in situ calibration standard. Since the
strongest Si peak (28.466° 2θ) overlaps with nephrite
peaks of interest, the second-strongest Si peak at 47.344°
2θ was used for calibration.
Results and discussion
Phase identification
In addition to XRD, RM is emerging as a valuable
analytical technique for in situ phase identification of
Chinese jades. Since different laser lines have been
reported for the characterization of nephrite (tremolite
and actinolite) and other amphibole minerals [9, 15, 18–
20], the spectral quality of Raman data collected at three
different excitation wavelengths on a late Neolithic/early
Shang blade (Zhang) were compared to determine
optimized collection conditions. In situ XRD analysis
shows that this jade is composed of ferroactinolite
(Ca2Fe5Si8O22(OH)2; JCPDS 23-0118) with minor impu-
Table 2 Early Chinese jades
from the collection of the Art
Institute of Chicago, arranged in
descending order from lightest
to darkest in color
Object description,
AIC accession no.
Attributed date Color Apparent composition
(Raman) (Fe/Fe+Mg)
(% p.f.u.)
Ring (huan), 1950.540 Neolithic period,
probably Liangzhu culture,
third millennium BCE
ivory white 0.37
Collared disc, 1950.240 Late Neolithic /
Shang dynasty,
second millenium BCE
very pale green,
mottled
0.87
Blade (zhang), 1950.318 Late Neolithic /
early Shang dynasty,
second millenium BCE
dark green,
almost black
29.73
Fig. 1 Modern reference geological samples of nephrite from China and neighboring regions
793
rities of chlorite (clinochlore; JCPDS 10-0183), and
ESEM/EDX shows it contains approximately 8% Mg,
5% Fe, 0.8% Al and 0.3% Mn. Although the Zhang blade
shows one of the highest fluorescence backgrounds of all
jades examined, it was possible to detect diagnostic
spectral features with all three laser excitations (Fig. 3).
However, based on further tests on other nephrite speci-
mens, the 532 nm laser was selected as the most effective
and versatile for successful collection of detailed spectra on
jades of various colors. Fluorescence was not completely
avoided, even using a 785 nm laser such as the one
integrated into the portable RM. However, 30-s collection
times allowed the acquisition of high-quality spectra on
most geological specimens with the portable instrument.
An exception was sample 7, which proved too fluorescent.
All of the diagnostic bands for nephrite are recorded
(Fig. 4), thus demonstrating the successful application of a
field-portable Raman spectrometer to Chinese jades for the
first time. Unfortunately, the lower resolutions of such
instruments limit the resolution of the bands in the region
200–300 cm−1 (the external lattice mode region, with
characteristic vibrations of O–H–O groups) and 300–
450 cm−1 (dominated by bands attributed to Si–O–Si and
O–Si–O bending modes and metal–oxygen bands), vibra-
tions that are characteristic of nephrite (tremolite–actino-
lite) and allow their clear distinction from other amphibole
minerals. Moreover, analysis with incident excitation light
of 785 nm precludes examination of the spectral region
centered at 3650 cm−1, which offers some useful features.
The availability of a benchtop spectrometer is always
advantageous, as it offers higher spectral resolution and
more flexibility in the selection of multiple excitation
wavelengths; however, portable RM carries an evident
advantage for the in situ examination of excavated jades at
archaeological sites, allowing fast and unambiguous dis-
crimination of nephrite jades from other stone artifacts.
XRD results for the geological specimens demonstrate a
shift from tremolite (JCPDS 44-1402) in samples 5 and 7 to
actinolite (JCPDS 41-1366) in samples 11 and 10 (Fig. 5).
Although peak intensities do not correspond well to
calculated values, peak positions show relatively good
agreement across the monoclinic spectrum. The lack of
agreement in relative intensity is not surprising, as
preferred orientation in bulk specimens is likely to
confound peak-matching programs. In addition to the
Fig. 2 Jade objects from the
collection of the Art Institute of
Chicago (1950.540; 1950.240;
1950.318; Bequest of Mrs
Edward—Louise B.—
Sonnenschein)
Fig. 3 Spectra obtained from a Zhang blade (nephrite, Late
Neolithic/early Shang dynasty, second millennium BCE) with
several excitation lines (no multipoint baseline correction was
applied to the data)
794
tremolite or actinolite spectra, there are a few unidentified
trace peaks. In the actinolite samples, peaks shift to lower
angles by as much as 0.2° 2θ. This is especially
pronounced in samples containing the highest Fe fraction
(sample 10). This is demonstrated in the enlarged view of
the 18 to 28° 2θ region in Fig. 6. Both of these observations
are consistent with the substitution of the larger Fe2+ cation
on Mg2+ sites, which increases the lattice parameter and
hence decreases the diffraction angle. Evans and Yang
verified this in a wide range of actinolite/tremolite single
Fig. 4 Comparison of spectra
recorded on nephrite geological
reference sample 11 with the
portable (a; λ0=785 nm) and
bench-top (b; λ0=532 nm)
micro-Raman spectrophotome-
ter (spectra have been normal-
ized with respect to the sym
Si–O–Si stretching band cen-
tered at 673 cm−1 to facilitate
comparison). The inset shows
details of the spectral features in
the O–H stretching region, de-
tectable only with 532 nm
excitation
Fig. 5 X-ray diffraction spectra
for samples 5 and 7, tremolite,
and samples 11 and 10,
actinolite
795
crystals [21]. Although the higher Fe content and light-to-
dark green color correlate with the actinolite spectra,
selective XRD peak analysis over longer counting times
would be required to quantitatively assess peak shift with
color.
Elemental composition and color
As shown in Table 1, XRF of the geological reference
specimens confirmed that the compositional variabilities of
these nephrites fell within a narrow range for most.
Exceptions are sample 3 from Taiwan, showing signifi-
cantly higher contents of Fe, Ni, Cr, and Mn with respect to
the other Chinese nephrites, and two samples from
Xiaomeling, Jiangsu province, showing unusually high
contents of Sr, thus easily separating them from the rest of
the group (sample 7 is shown as representative example in
Table 1). In agreement with data on New Zealand nephrite
[10], data collected for this study indicate that cationic
content does not easily correlate with the perceived color of
the jades.
For example, if one follows the variation in color with Fe
content, CIELAB coordinates calculated for the visible
spectra recorded on differently colored jades in order to
acquire numerical information on color [22–24] showed
sample 10 (brownish green) to have the lowest L* value,
even though its Fe content is approximately three times
lower than sample 3 (olive green) (Table 1). Almost
identically low values of L* were measured for samples 11
(yellow-green) and 3 (olive green), even though the latter
has approximately thirteen times the Fe content of the
former. High variability was observed in particular for the
low-iron nephrites, whose colors vary from white to
yellow-green in a nonmonotonic fashion with increasing
Fe content. For example, samples 5 (white) and 11 (yellow
green) have comparable total Fe contents, with the white
nephrite showing a higher Fe/Fe+ Mg (% p.f.u.) ratio.
Thus, although previous research has suggested that
increasing Fe content in nephrite is correlated with a
darkening of the color [1], a simple linear correlation does
not apply.
Visible reflectance spectra recorded on the geological
specimens are consistent with extensive discussions in the
literature on amphibole minerals [25, 26]. Two dominant
processes are responsible for the spectral features observed
in the electronic spectra of nephrite: metal to ligand charge
transfer (MLCT: Fe2+→O) involving Fe2+ in M1, M2, M3
andM4 sites and Fe3+ andCr3+ inM2,M1 andM3 sites; and
homonuclear (Fe2+→Fe3+) or heteronuclear (Fe2+→Ti4+)
intervalence charge transfer (IVCT). Most samples
analyzed for this study clearly showed the tail of the
MLCT Fe2+→O band, a band that has its absorption
maximum in the higher energy UV region (Fig. 7a). All of
the samples also exhibited a decrease in the reflectance
factor at around 700 nm, related to Fe3+ spectral features.
Only sample 3 showed weak absorption bands at 630, 658
and 691 nm, due to the presence of Cr3+(Fig. 7b), detected
by XRF in relatively high abundance solely for this
sample. Sample 3 also showed a small absorption at
420 nm due to a Fe3+ crystal field (CF) transition, with
other weak absorptions at 400–600 nm associated with
spin-forbidden bands of Fe2+. With the exception of
sample 5, which is white in color, and has a maximum
Fig. 6 Enlarged view of X-ray
diffraction of actinolite samples
2 and 10, shown to demonstrate
the peak shifts with increasing
Fe content
796
reflectance factor (λmax) at 459 nm, all of the other
samples owe their green color to a λmax in the green
portion of the visible spectrum between 500 and 530 nm.
Previous research has suggested that relatively high
chromium content in nephrite can be detected by a “strong
intensity in the Raman spectra in the 6000 cm−1 range”
([8], p. 34); however, this study did not confirm such a
linear influence of the presence of Cr in the high
wavenumber region of the Raman spectra. In fact, with
the exception of sample 3, characterized by significant Cr
content, all other geological reference samples examined
had only trace amounts of Cr (reported in [14] at 0.05–
0.07%). Nevertheless, Raman spectra for samples 10 and 7
also showed an increasing slope in the baseline at
wavenumbers higher than 4000 cm−1. This spectral feature
seems rather to correlate with higher reflectance in the red
region of the visible portion of the spectrum. Those
samples (2, 5, 9, 11) that, on the contrary, show an
absorption in such region have flat baselines at wavenum-
bers higher than 4000 cm−1 (Figs. 7 and 8). These findings
underscore the need for a more critical, less simplified
approach towards the assessment of the complex interrela-
tion between color, cation composition, and broad features
in the Raman spectra.
Fig. 7 a,b a Visible reflectance
spectra of geological nephrite
samples: a 10; b 3; c 5; d 9; e 2.
b Detail of the visible reflec-
tance spectrum of nephrite
sample 3, showing characteristic
spectral features linked to the
presence of Fe2+, Fe3+ and Cr3+
ions
Fig. 8 Micro-Raman spectra of
nephrite samples 3 (c) and 10
(b) showing a sloping, increas-
ing baseline feature at wave-
numbers higher than 4000 cm−1
and sample 9 (a), showing a
lack thereof (λ0=532 nm)
797
Structural implications for the determination
of geological origin
The crystal structure of nephrite consists of a double-chain
silicate with brucite-like strips sandwiched between
opposed silicate chains with linking cations of nearly
octahedral coordination [15]. A recent study by RM [15]
demonstrates that analysis of the fine structure associated
with the O–H stretching Raman bands of nephrite allows
the cationic substitution of Fe (or, more precisely, Fe+Mn)
to be determined for Mg in M1 and M3 sites. This allows a
reasonable approximation of the Fe content in nephrite to
be made, which, in turn, can be correlated to the geological
origin of the material: from nephrite deposits associated
with magnesian marbles or with serpentinized ultramafic
rocks [1]. Fe/Fe+Mg ratios (% p.f.u.: XR% in the
following) of between 0.4 and 7.4 can be ascribed to
magnesian marbles (dominant in mainland China), and
values >8.1 to serpentinized ultramafic rocks (more
common in other regions worldwide). In the present
study, only samples with XR% >10 showed significant
splitting of the O–H stretching band (centered at 3675 cm−1
and the only one observed when Mg is the sole occupant of
the M1 and M3 sites), with a medium peak at 3664 cm−1
(indicative of MgMgFe or MgFeMg occupancy) and
emergence of the peak at 3645 cm−1, indicative of FeFeMg
or FeMgFe substitution (observed as a weak but well-
defined peak only for sample 3 and for the dark-colored
Zhang blade). At XR% >3, one clearly observes a small
peak at 3662 cm−1, while at XR%∼1 only a very weak peak
at this wavenumber is discernable (Fig. 9).
Heat treatment of nephrite
RM also proved invaluable for the noninvasive detection of
instances of heat treatment in nephrite, a phenomenon
previously studied with XRD and Fourier transform
infrared spectroscopy (FTIR) on samples removed from
the jades [17]. Artificial jade treatment is of great interest to
archaeologists and art historians studying Chinese jades,
because its detection and identification allows a deeper
understanding of the culture that produced and used such
objects. With regards to heating, two principal theories
have been proposed: some jades may have been heat-
treated to soften the stones prior to working them;
alternatively, jades may have been subjected to heat during
ritual ceremonies. Additionally, detection of heat treatment
is of importance because forgers are thought to treat
modern jade objects with heat in order to simulate the
whitened, altered appearance of ancient jades.
The dehydration process of tremolite/actinolite as well
as the colors observed following heating have been
previously described by Douglas [17]. The color changes
exhibited by the low-iron nephrite pebble used in this
study, from translucent grayish yellow green to opaque
white with increasing temperature (Fig. 10), are connected
to a solid-state reaction where diopside and cristobalite
pseudomorphs of tremolite crystals are generated following
the reaction:
Ca2Mg5Si8O22 OHð Þ2
!

 2CaSiO3  5MgSiO3 þ SiO2 þ H2O
nephrite tremoliteð Þ pyroxene diopsideð Þ cristobalite water
Fig. 9 Micro Raman spectra of (a) blackish Zhang blade, (b) olive
green nephrite, sample 3, (c) brownish green nephrite, sample 10,
(d) light green nephrite, sample 9, (e) greenish yellow nephrite,
sample 11 (λ0=532 nm; after multipoint baseline subtraction). For
ease of reference, each respective Fe/Fe+Mg (% p.f.u.) ratio derived
from the normalized intensity of the Raman band in this region is
given
798
Some darkening was observed at the edges of the
nephrite specimens heated from 500 °C to 800 °C,
successively turning from a reddish brown color at
900 °C to an orange-red color at 1100 °C. In the
corresponding area of the unheated nephrite sample, the
material has a slight greenish color, possibly representing a
relatively Fe-rich area.
Because of its high spatial resolution and ability to detect
mineral oxides, RM allowed the characterization of the
origin of the blackish and reddish color that develops at the
uppermost surfaces of the samples and whose nature was
previously only hypothesized. As illustrated in Fig. 11, the
present analysis shows that finely dispersed carbonaceous
particles, together with localized accumulation of Fe
oxidation products (hematite, magnetite and possibly
maghemite or other Fe oxides characterized by a band at
700 cm−1) are responsible for the brownish color of the
slices heated between 600 °C and 800 °C [27]. Only
hematite and magnetite are detected at 900 °C and at
1000 °C, while at 1100 °C the diagnostic peaks for
hematite and diopside are detected together, suggesting that
these two minerals are intimately mixed.
With respect to changes affecting the bulk of the heated
samples, ten to fifteen areas were probed on each slice
representative of a discrete temperature step, and the
spectra collected confirmed that no structural or chemical
change is detectable for temperatures up to 900 °C: in fact,
only the characteristic bands of nephrite are evident (with
bands at 129 m, 162 m, 224 m, 251 vw, 266 vw, 302 vw,
351 vw, 369 w, 393 m, 416 w, 437 vw, 526 w, 673 vvs,
930 m, 948 w, sh, 1029 m, 1060 s, 3662 vw, 3676 vs cm−1).
Starting at 900 °C, the bands of diopside (198 w, 233 w,
329 m,395 m, 560 vw, 668 s, 1013 vs cm−1) are detected,
sometimes together with those of nephrite, as illustrated in
Fig. 12, clearly indicating that structural and chemical
change is either in its inception or in full progress or
completed at this temperature, depending on the specific
area probed. In some areas of the specimen heated up to
900 °C, nephrite is the most predominant phase, in other
areas it is diopside. In the O–H stretching region, the band
at 3675 cm−1 decreases in intensity and broadens as a result
of oxidation of Fe ions present in the lattice, while the weak
band at 3661 cm−1 disappears (disappearing at 800 °C).
Finally, at 1000 °C and 1100 °C no spectral feature is
detected in the O–H stretching region, confirming that the
only phase remaining is diopside.
The RM study of heating described here corroborates
previously published results obtained with FTIR on samples
removed from the same specimens and analyzed in a
diamond microcompression cell [17]. However, RM
benefits from higher spatial resolution (over one order of
magnitude lower, as microsampling affects areas of the
order of millimeters) and, being noninvasive, allows
thorough mapping of the entirety of the objects examined.
The detection of isolated crystals of diopside in the
uppermost areas of the nephrite pebble heated as low as
Fig. 10 Slices of a low-iron
nephrite pebble from Hetian
(Khotan), Xinjiang province,
heated individually to increasing
temperatures up to 1100 °C (the
largest slice measured
approximately 3 cm)
799
600 °C (an inhomogeneity in the nephrite) demonstrates that
caution should be exercised when analysis has to rely on the
removal of only one small sample from a determinate area.
Conclusions
This study presents an innovative, totally noninvasive
approach to the study of Chinese jades, using complemen-
tary analytical methods to not only identify nephrite but
also to characterize important parameters related to color,
geological origin and natural or human-induced alterations.
Visible spectrophotometry provided accurate measure-
ments of color, although the complex electronic spectra
obtained for the different geological samples reflect the
extreme variability in the color and shade of nephrite, and
their detailed interpretation will require further study. The
techniques of RM, XRD, and XRF are rapid and effective
means to provide in situ jade characterization, and together
they provide data on mineralogical identification, elemen-
tal composition, and crystallographic structure. In partic-
ular, RM has great potential as it allows selective study of
the surface structure of jade at the microscopic level in
order to identify features such as heating, accretions, color
shading, and texture. Although resolution is limited,
portable RM is particularly useful for field applications,
allowing for rapid discrimination of nephrite from jadeite
(not documented in the context of Chinese jade artifacts
before the eighteenth century) and other jade-like stone
materials such as quartz, calcite, serpentine, chlorite, and
talc. The use of RM to determine Fe/Fe+Mg ratios (% p.f.u.)
in jades has proven useful as a method of classifying
artifacts and gaining insights into their geological origins.
Fig. 11 Micro-Raman spectra of the oxidation products of Fe
compounds conferring a characteristic reddish brown color to the
uppermost surface of the heated jades (λ0=532 nm; after multipoint
baseline subtraction). At 1100 °C only diopside and hematite are
evident, at 900 °C hematite and magnetite, at 800 °C hematite,
magnetite and maghemite. In the inset, a microphotograph acquired
through the integrated microscope of the Raman spectrometer shows
the particles of iron oxide interdispersed in the nephrite/diopside
matrix in the heated samples (the bar measures 250 μm)
Fig. 12 Raman spectra excited
with 532 nm laser on the slices
of low-iron nephrite pebble
heated in discrete steps from 0
to 1100 °C, illustrating chemical
and structural changes asso-
ciated with heating. At top,
characteristic Raman bands for
nephrite are marked, at bottom,
the characteristic bands for
diopside. At right, the O–H
stretching region is shown, with
two curves for the sample
heated to 900 °C to illustrate the
presence of either nephrite or
diopside in the different areas
probed. (All spectra shown after
multipoint baseline subtraction)
800
The accumulation of a database of Fe/Fe+Mg (% p.f.u.)
ratios from scientifically excavated archaeological objects
and well-documented museum objects will prove invalu-
able for the ultimate provenancing of undocumented
Chinese jades.
Acknowledgements Conservation science research at the Art
Institute of Chicago is made possible by a generous grant from the A.
W. Mellon Foundation. The Community Associates of the Art
Institute are thanked for their support of the microfocus XRF. Elinor
Pearlstein and Jay Xu, curators of Asian Art at the Art Institute of
Chicago, are also thanked. The assistance of Susie James, Alyson
Whitney, Jerry Carsello and Benjamin Meyers for data collection at
Northwestern is gratefully acknowledged. X-ray facilities at NU are
supported by the MRSEC program of the National Science
Foundation (DMR-0076097) of the Materials Research Center of
Northwestern University. Wen Guang, Feng Min, Tan Li-ping and
Joseph Hotung are thanked for providing reference nephrite samples.
References
1. Wen G, Jing Z (1996) Acta Geol Taiwan 32:55–83
2. Douglas JG (2005) A review on some recent research on early
Chinese jades. In: Scientific examination of art: modern
techniques in conservation and analysis, from the Arthur M.
Sackler Colloquia of the National Academy of Sciences. The
National Academies Press, Washington, DC, pp 206–214
3. Wen G (1997) Colloq Art Archaeol Asia 18:105–122
4. Wen G (1998) In: Tang C (ed) East Asian jades: symbol of
excellence, vol. 2. Chinese University of Hong Kong, Hong
Kong, pp 217–221
5. Wen G, Jing Z (1992) Geoarchaeology 7:251–255
6. So J, Douglas JG (1998) Understanding and identifying jades
from the Hongshan Culture. In: East Asian jades: symbol of
excellence, vol. 1. The Chinese University of Hong Kong,
Hong Kong, pp 148–163
7. Casadio F, Xu J, Pearlstein, Faber KT, Knowles A, Jing Z
(2006) A stone kneeling figure in the Art Institute of Chicago:
new evidence from scientific investigations and archaeological
finds in China. In: Douglas JG, Jett P, Winter J (eds) Scientific
Research on the Sculptural Arts of Asia, Proceedings of the
Third Forbes Symposium at the Freer Gallery of Art. Archetype
Publ., London (in association with the Freer Gallery of Art,
Smithsonian Institution, Washington, DC) (in press)
8. Xu JN, Huang E, Chen CH, Tan LP, Yu BS (1996) Acta Geol
Taiwan (Special Issue on the Mineralogical Study of Archaic
Jades) 32:11–42
9. Shurvell HF, Rintoul L, Fredericks PM (2001) Int J Vibr Spec 5
(5):4 (see http://www.ijvs.com)
10. Wilkins CJ, Tennant CW, Williamson BE, McCammon CA
(2003) Am Mineral 88:1336–1344
11. Gendron F, Smith DC, Gendron-Badou A (2002) J Archaeol
Sci 29:837–851
12. Smith DC, Gendron F (1997) J Raman Spectrosc 28:
731–738
13. Douglas JG (1996) Acta Geol Taiwan (Special Issue on the
Mineralogical Study of Archaic Jades) 32:43–54
14. Douglas JG (2003) Exploring issues of geological source for
jade worked by ancient Chinese cultures with the aid of X-ray
fluorescence spectroscopy. In: Jett P, Douglas JG, McCarthy
B, Winter J (eds) Scientific Research in the Field of Asian
Art, Proceedings of the First Forbes Symposium at the Freer
Gallery of Art. Archetype Publ., London (in association with
the Freer Gallery of Art, Smithsonian Institution, Washington,
DC), pp 192–199
15. Chen TH, Calligaro T, Pagès-Camagna S, Menu M (2004) Appl
Phys A 79:177–180
16. Middleton A, Ambers J (2006) Case study: Analysis of
nephrite jade using Raman microscopy and X-ray fluores-
cence spectroscopy. In: Edwards HGM, Chalmers JM (eds)
Raman spectroscopy in archaeology and art history. Royal
Society of Chemistry Analytical Spectroscopy Monographs,
Cambridge, UK, 23:403–411
17. Douglas JG (2001) The effect of heat on nephrite and detection
of heated Chinese jades by X-ray diffraction (XRD) and
Fourier-transform infrared spectroscopy (FTIR). In: Proc Conf
Archaic Jades Across the Taiwan Straits. Taipei, Taiwan,
pp 543–554
18. Blaha JJ, Rosasco GJ (1978) Anal Chem 50(7):892–896
19. Lewis IR, Chaffin NC, Gunter ME, Griffiths PR (1996)
Spectrochim Acta Part A 52:315–328
20. Rinaudo C, Belluso E, Gastaldi D (2004) Miner Mag 68 (3):
455–465
21. Evans BW, Yang H (1998) Am Mineral 83:458–475
22. Berns RS (2000) Billmeyer and Saltzman’s principles of color
technology, 3rd edn. Wiley, New York
23. International Commission on Illumination, Bureau Central de
la CIE (1931) Proc 8th Session. Cambridge, England,
pp 19–29
24. CIE (1986) Colorimetry, 2nd edn (No. 15.2). Commission
Internationale de L’Éclairage, Vienna, Austria
25. Karr C Jr (ed) (1975) Infrared and Raman spectroscopy of lunar
and terrestrial minerals. Academic, New York
26. Hunt GR, Salisbury JW, Lenhoff CJ (1973) Mod Geol 4:85-106
27. de Faria DLA, Silva SV, de Olivera MT (1997) J Raman
Spectrosc 28:873–878
801