VNU Journal of Science, Earth Sciences 26 (2010) 121-127

Objective and Subjective Factors Influence on
Demand of Drainage by Pumping in Red River Delta
Dang Ngoc Hanh*
Vietnam Academy for Water Resources
Received 22 October 2010; received in revised form 19 November 2010
Abstract. The demand for newly constructing of drainage pumping station in Red river delta has
increased recently. Total drainage volume for the whole region in 2006 was 2406.8m3/s, and it was
predicted to increase to 5181.3m3/s in 2020 [1]. The average drainage coefficient for the year from
2010 to 2020 will be 7.0 l/s per hectar, three and a haft time higher than the average drainage
coefficient for the period from 1954 to 1973; 1.8 times higher than 1973 to 1976 and 1.3 times
higher than 1976 to 2000. This article aims to provide the primary analysis of objective (drainage
requirement) and subjective factors (socio-economic condition, psychology, information, etc)
influencing on the demand of drainage. The out come shows that the draingae demand seems to be
impacted by subjective factors rather than objective ones.

1. Introduction∗

Resources Planning shows that the drainage
requirement area for 2010 based on sustainable
senerior is 1,116,559 hectar [1]. Of which
drainage by pump serves for 731,432 hectars
(taking approximatly 63%) and gravity drainage
takes 435,127 hectars (taking 37%).

Since 1954 Government had paid a lot of
attention on drainage in Red river delta. This
attention has increased recently. The scale of
drainage struture very much depends on
drainage coefficient which is calculated based

on drainage unit on area unit (l/s-ha). Through
water resources planning, the drainage
coefficient has been increased, from 1954 to
1973, the drainage ratio in region was 2.03l/sha; from 1973 to 1976 it was 3.89 l/s-ha; and
from 1976 to 2000 it was 5.32 l/s-ha; During
the period from 2010 to 2020, the drainage
coefficient was predicted from 6.88 l/s-ha to
6.90 l/s-ha. There are a number of drainage
pumping stations which were designed at the
drainage coefficient of 12.50 l/s-ha.

Based on the drainage requirement,
Ministry of Agriculture and Rural Development
(MARD) issued the list of investment for the
duration from 2011 to 2015, document
3505/BNN-KH on 28 October 2009, which
included thirty six projects with the total
proposed investment of 14,043 billion VND.
Among this list, there were 29 projects on
drainage. Regarding new construction of
pumping stations, within sixteen projects there
were only two irrigation pumping station
projects, fourteen projects were constructing of
drainage pumping station. Total investment for
newly constructing of sixteen pumping station
was 5,425 billion VND, of which 5,105 billion
VND was the cost for fourteen drainage
pumping station projects (taking 94%). This

In regards to drainage service area, the

master plan developed by Institute for Water

_______
∗

Tel.: 84-4-38522293.
E-mail:

121


122

D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127

number shows the necessity of new
construction of drainage pumping station in
near future, as well as indicates the important
role of drainage in developing the irrigation and
drainage system in Red river delta.
In order to investigate the main reason for
increasing in the requirement for more drainage
pumping station, this article will examine the
objective and subjective factors which
impacting on drainage requirement as well as
find out the appropriate attitude for drainage.
2. Objective factors
Objective factor impacting on drainage
requirement is intensive rainfall. Through


analyzing the changes in rainfall statistics, the
objective factors on drainage and drainage
solution will be examined. The analysis of
pumping drainage has been studied in 7
meteorology stations including Hai Duong,
Hung Yen, Ha Dong, Ha Nam, Nam Dinh,
Ninh Binh and Thai Binh. The statistical daily
rainfall data from 1976 to 2008 [2] for Phu Ly
station has been used for illustrating the 1, 3, 5,
and 7 days of the maximum intensive rainfall
and the total rainfall at those stations (figure 1).
The rainfall data for other stations in Red river
delta is also demonstrated, and the evaluation as
below

Figure 1. Example of the 1, 3, 5, and 7 days maximum intensive rainfall
and total rainfall at Phu ly meteorolofy station.


D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127

123

- One day maximum intensive rainfall
seems to reduce, especially in Nam Dinh
meteorology station the rainfall has been
avaragely decrease 3.5 mm/year during 33
years of recording. Others seven meteorology
stations also have the same trend as Nam Dinh
station. The statistical data for Ha Dong

metrorology station shows a particular point of
intensive rainfall in November 2008 due to over
intensive rainfall on that year.

- The total rainfall which was measured at
five among sevem meteorology stations seems
to decrease. The most reduction can be seen in
Hung Yen station in about 12mm/year. Nam
Dinh, Thai Binh, Phu Ly and Ninh Binh
stations have the reduction from 9 mm/ year to
10 mm/year. The figure of Ha Duong seems to
be stable. Ha Dong station has a trend of
increasing but this trend influenced by
historical rainfall in November 2008.

- In regards to three days maximum
intensive
rainfall,
five
among seven
meteorology stations including Nam Dinh, Thai
Binh, Ninh Binh, Phu Ly and Hung Yen seem
to have a reduction in rainfall from 0.6 mm to
3.7 mm/year. There is only Hai Duong station
showing the increase in three day maximum
intensive rainfall, but it is negligible, only 0.6
mm/year. Regarding Ha Dong station, althought
it has been showed the increasing trend, it was
just an impact of historical rainfall in November
2008. If the figure is only presented for Ha

Dong station by 2007, it also shows the
reduction trend as same as other meteorology
stations.

- In regards to characteristic of intensive
rainfall, it seems almost 1, 3, 5, or 7 days of
intensive rainfall often happened in the long
period of rain. This issue brings the nagative
impact on rainfall model simulation.

- Regarding five day maximum intensive
rainfall, five within seven meteorology stations
show the reduction trend. The remarkable
reduction can be seen in Nam Dinh and Hung
Yen meteorology station, decreasing from 3.6
to 3.7 mm/year. There was no change in Hai
Duong meteorology station. The figure of Ha
Dong station seems to have the increasing trend
but it was because of historical rainfall in
November 2010
- Concerning seven day maximum intensive
rainfall, five among seven meteorology stations
show the reduction trend of about 3 mm/year.
The figure of Ha Duong station shows no
changes. The figure of Ha Dong station seems
to have the increasing trend but it was because
of historical rainfall in November 2010

Analyzing the rainfall changes and trend
shows that the comparison of increasing in

drainage coefficient and drainage rainfall seems
to contradict. This contradiction can be
explaned as following
- Previously, the TCVN 285-2002 and
TCVN 5090 and other regulations regulated
that the drainage capacity in responding for
rainfall frequency of about 10 to 20%. Due to
difficulties in economy, the drainage capacity
could be selected at the rainfall frequency of
12%, 15% and event 20%. Nowadays, almost
of all planning and design often based on the
rainfall frequency of 10%.
- Design drainage coefficient also depends
on drainage model. In the past, because of
economic condition, the planner could be based
on rainfall model with fewer disadvantages in
order to reduce the design drainage coefficient.
Analyzing the changes in drainage of
intensive rainfall and factors impact on drainage
requirement and shows that rainfall seems not
to be an objective factor. Socio-economic,
changing in cropping pattern or other factors
might be the subjective factors impacting on
drainage requirement.


124

D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127


In
addition,
due
to
topography
characteristics of Red river basin slopes from
Northwest to Southeast, particularly the delta is
in the pan shape topography with the higher
elevation land located along the river bank and
deep valley is located on cultivated farm, these
conditons have nagative impacted on drainage
actitities, especially the central delta where
drainage by tide has been not in practice.
Therefore, the drainage pumping stations are
often located in the Red river delta such as
North Nam Ha, and Southwest of Nhue river
system.

will remarkably increase which pressing on the
drainage demand. Particularly, it was predicted
that during the period from 2010 to 2020 and
2030, rainfall in Red river delta region will
increase by 1.6% to 2.3% in comparison to the
time from 1980 to 1999.

3. Prediction factors on climate change
impact on drainage solutions

- The combination of rainfall and sea level
rise will narrow down the gravity drainage area

in coastal zone in the North. A large area will
be inundated and semi-inundated. According to
Prof Dr Dao Xuan Hoc [4] pointed out that the
inundated area in Red river delta will increase
to 550,000 hectares, 650,000 hecrares if sea
level rises up to 0.69 m and 1.00 m
respectively. In addition, the river level will rise
at the average level from 0.5 m to 1.0 m, exceed
alert 3. That means the water level in river
nearly approaches high crest of current dike. If
sea level rises up to 0.69 m, the area in Red
river delta with the elevation is below 0.8 m
(133,221 hectares) will be inundated, the area
with the elevation is higher than 2.2 m (300,319
hectares) will be semi-inundated; If sea level
rises up to 1m, the complete inundated area
below 1.5 m will be 181,917 hectares, and
semi-inundated area with the elevation below
2.5 m will be 365,431 hectares.

There are two main drainage solutions in
Red river delta which including pumping and
gravity. Gravity drainage takes small
percentage (below 30%) concentrating on
coastal zone areas. This drainage solution
makes used of ebbing tide for drainage. In
which, among 297,600 hectares of full gravity
drainage in the region [3], there are 279,300
hectares (taking 94%), located along coastal
zone in South and North of Thai Binh province,

central and South of Nam Dinh province, as
well as Hai Phong.
Climate change scenario, and sea water rise
for Vietnam has been developed by Ministry of
Natural Resources and Environment which was
declared in June 2009 indicating:
- According to low emission scenario,
rainfall in Red river delta and Thai Binh river
basin will increase 5%, while rainfall from
March to May will reduce from 3% to 6%. In
medium greenhouse gas emission, rainfall in
Red river delta will increase up to 10%, and
rainfall from March to May will decrease from
6% to 9%. Therefore, rainfall in rainy season

- In regards to low emission, medium
emission (B2) and high emission scenario
(A1F1) , sea level rise will increase by 0.65 m,
0.75 m and 1.0 m respectively at the end of this
century. From 2020 to 2030, it was predicted ,
sea level is projected increasing from 12 to 17
cm.

Climate change and sea level rise in any
scenerios always bring the strong impact on
drainage solutions, therefore drainage by pump
seems to be the only one solution for Red river
delta in the future



D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127

4. Impact of socio-economic development on
drainage requirement
Economic development is related to the
process of development, construction, upgrade,
and comprehension of infrastructure and the
changing of land use. Base on the research of
rainfall-runoff, drainage is affected by surface
area.
The research of rainfall- runoff is showed
the lost volume on the surface of hydraulic
structure by waterproof material which is tiny.
The rate of rigid surface increase which is
opposite with the loss of water therefore it is
danger of the amount of drained water and
describing by run-off coefficient.

125

The conception of run-off coefficient is
calculated by the ratio between suface water
(mm) by rainfall and the amount of water (mm)
σRoff = Y/P, where: Y: run-off by rainfall; P:
precipitation
In the developed countries, run-off
coefficient is researched by the sceintiest of
irrigation and drainage. After that, it will be
determined the other surface factors in different
areas, different regions, different crops.

Meanwhile, the research in run-off coefficient
in Vietnam which is too limited, even in
textbook, guide book the run-off coefficient is
copied from abroad research results.

Run-off coefficients of surface
Surface classification
Grass
- Sand, steep 2%
- Sandy loam, steep 2%
Urban land
Rural land
Stone
Streetside
Roof

Run-off coefficients σRoff
0.05-0.10
0.13-0.17
0.70-0.95
0.50-0.70
0.70-0.85
0.75-0.85
0.75-0.95

Surface classification
Industrial zone
- density rarely
- density dense
Amusement park

Railway
Concrete
Mặt bê tông

Run-off coefficients σRoff
0.50-0.80
0.60-0.90
0.10-0.25
0.20-0.35
0.20-0.40
0.70-0.95
0.80-0.95

Source: AFTER CHOW, 1962

Obviously, run-off coefficient is on surface
by infrastructure is compared with the other
land use objects, especially over 90% rainfall
will be a run-off on surface of cement, roof in
case no water storage or other multi use which
will have to drain.
In this case, the statistic of transportation in
Red river delta show the density of national,
provincial, districial highways is high, in range
0.5 km/km2 [1], if the wide of road is 20m, the
area will has 1% of total zone. Similarity, the
density of village road is 1.81 km/km2 [1], if the
wide of road is 3m, the area will has 0.54% of
total zone. The sup up of road will grow up to


1.54% (equal to 15,718 hectares). The area for
transportation is huge, the plan for land use in 6
provinces in the central delta includes Bac
Ninh, Ha Tay, Hai Duong, Hung Yen, Ha Nam
and Nam Dinh [5] in 2005, the transportation
area has 48,619 hectares, plan to 56,218
hectares in 2010, increase 7,599 hectares.
In the other, economic development will be
incresingly land for industry used. Statistical
data for Industry land used at 2005 in six
provinces in central of the delta and Thai Binh
province is 8,282 hectares, planning for 2010
will be predicted to 25,098 hectares, more than
three time higher. The change of land used


126

D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127

mostly form cultivation land and pond or lake
land, that is one of big reasons to incresingly of
darinage
requirement
because
ron-off
coefficient for the industry land used from 0.5
to 0.9, in other for the cultivation land run-off
coefficient is only from 0.3 to 0.6 [6].
Urban land increased dramatically in 2005,

7 provinces in central delta has 8109 hectares
urban land which would be increased to 14,290
hectares area in 2010, 1.76 times increasing.
The cultivated land has been transferred into
urban land, which is the main reason causing
the demanding on drainage.
Rural land also increase many times which
is also transferred from cultivated land. In 2005,
7 provinces had 69,996ha, planned to 74,748
hectares in 2010, increasing of 4752 hectares.
Therefore the drainage demand will increase by
run-off coefficient from 0.5 to 0.7 in rural land
which reference from aboard document (note
that the population density in Red river delta is
122,000 people in a square kilometre, may be
higher than foreign countryside). It is much
higher than the run-off coefficient in cultivated
land [6].
The increasing in non cultivated land has
been contradicted is opposited with the
reduction area of lake, pond, stream and river.
In 2005, 7 provinces in central delta had 61,482
hectares; however its plan has gone down to
58,064 hectares in 2010. Almost of 3418
hectares reduction is pond and lake which are
water storage area and can be use to regulate
rainfall, because the using purpose might be not
changed from natural river and stream. There is
information show that, In Hanoi capital city,
80% of water surface area has regulated

capacity which has been leveling for
construction for 50 years by a source.
In addition, the area for rice (which can
regulate drain water) in some areas in central

delta has dramatical fall to 386,641ha, 71,170ha
mitigration which is significant number. If the
rice area in Red river delta drecrese 40,700ha
from 2000 to 2005 (report of MARD) and in 10
years, 2000 to 2010, there is at least 111,870ha
rice area (10%) total natural area in this delta
change to non agricultural purpose. The change
of land use to non agricultural purpose is
leading to the increase of drainage demand
many times which is compared with rice land
and cultivated land.

5. Discussion and recommenation
Obviously, all socio-economic indicators
have been indicating land use planning in all
provinces in Red river delta showing the
objective fators for increasing the drainage
demand. In the above analysis shows the rise of
amount of drain, and the drainage capacity also
increases by the drainage demand for
incultivated land which is drained by day. There
are 2 factors to increase the domain of area and
which are the main causes in order to the
increase of drainage coefficient demand.
Throught out the analysis, again we are able to

realize that the cause of increase drainage
coefficient is by the subject factors than
objective factors.
Beside the above analysis, there may be an
impact which is media commucation. This
impact influences in decision making process of
decision makers. For example, the information
is transferred quickly via media people and
often does not purely reflect the actual
situations. This strongly impacts on making the
decision which consequently effects on
planning of drainage system. The design
parameters of the drainage system are affected
by a series of factors therefore the consultants
seem to choose the negative factors for


D.N. Hanh / VNU Journal of Science, Earth Sciences 26 (2010) 121-127

designing which enable for easy approval. In
order to mitigate these potential impacts, the
government will need to have regular
researches on monitoring and evaluation of
drainage system. Findings from these
researches will be the foundation for developing
appropriate drainage system in order to
minimize the investment cost and wasting rain
water.
The analysis also show that the pumping
drainage structures seem having bigger capacity

which might increase by 2 times, reaching the
capacity of 5181.3 m3/s in coming 10 years.
Hence, it is necessary to conduct the researches
on appropriate solutions in order to upgrade and
modernize the drainage pumping system in Red
river delta. Especially, there is a particular need
to find the proper solutions for on farm
drainage, effective use of rainwater, against
groundwater depletion and analyse effective
investment to prove to the society the effect and
necessity of investing on drainage structures.

127

References
[1] Institute of Water Resources Planning: The
report of drainage planning, the project of
planning for using water with multi objectives in
Red – Thai Binh river delta, code 5390
QD/bNN-KH, Hanoi, 2006.
[2] Hydrometeorological data Centre: The table of
statistical rainfall data at all stations in Hai
Duong, Hung Yen, Ha Dong, Ha Nam, Nam
Dinh, Ninh Binh, Thai Binh provinces from 1976
to 2008.
[3] Bui Nam Sach, Fundamental research in science
and practice of determine the drainage zone in
north delta. The thesis of master of science.
Water Resources University. Hanoi 2000.
[4] Dao Xuan Hoc, The plan of climate change

adaptation in agriculture and rural development
– Workshop Vietnam adapt with climate change,
31 July 2009 in Hoi An – Quang Nam
[5] Governmental resolutions in 2007 to adjust Land
use planning to 2010 and the land use planning
for 5 years in Hai Duong, Hung Yen, Ha Dong,
Ha Nam, Nam Dinh, Ninh Binh, Thai Binh
provinces
[6] Pham Ngoc Hai and NNK, the Textbook of
Planning and Design the hydraulic construction
system, Construction Publishing house, 2006.


VNU Journal of Science, Earth Sciences 26 (2010) 32-41

On some controversially-discussed Raman and IR bands
of beryl
Le Thi Thu Huong1, Tobias Häger2
1

Faculty of Geology, Hanoi University of Science, VNU, 334 Nguyen Trai, Hanoi, Vietnam
2
Institute of Geology, Johannes Gutenberg – University (Mainz, Germany)
Received 14 September 2010; received in revised form 28 October 2010

Abstract. Natural and synthetic beryl, Al2Be3Si6O18, from various deposits and manufacturers
were investigated with Raman, IR spectroscopy, Laser Ablation Inductively Coupled Plasma Mass
Spectrometry (LA-ICP-MS) and Electron Microprobe Analysis (EMPA). The Raman-band at
1067-1072 cm-1 and the IR-band at 1071-1207 cm-1 have been assigned till now either to Si-O or
to Be-O by different studies. Following the findings in this study that the position and Full Width

at Half Maximum (FWHM) of these bands were related to the concentration of silicon but not that
of beryllium, it stated that these bands were generated by the vibration of Si-O.
Keywords: Raman, Infrared spectroscopy, FWHM, band position, beryl.

1. Introduction∗

such as Raman and Infrared (Nasdala et al.
2004 [4]). As calculated by Kim et a. (1995) [3]
and described by Moroz et al (2000) [5], the
Raman band at 1067-1072 cm-1 has been
assigned to Be-O vibration. However, this band
was attributed by Adams & Gardner [1] and
mentioned in the study of Charoy et al. (1996)
[6] to the Si-O bond. Similarly, the IR band at
1071-1207 cm-1 has been assigned to Be-O
vibration by Plyusnina [7], Plyusnina &
Surzhanskaya [8] and to Si-O vibration by
Aurisicchio et al. [9], Manier-Glavinaz et al.
[10], Hofmeister et al. [2], Adams & Gardner
[1], Gervais & Pirou [11]. According to our
study, the features of both Raman and IR bands
(band position and band width) were clearly
related to the concentration of Si in the samples.
The band width was shown to be broader in the
samples containing a lower amount of silicon;

In this study we focused on one Raman
band at about 1067-1072 cm-1 and one IR band
at about 1071-1207 cm-1 of the cyclo-silicate
mineral beryl, Al2Be3Si6O18 (SiO2-67 wt%,

Al2O3-18,9 wt%, BeO-14,1 wt%, theoretically).
The study aimed to obtain a better
understanding of vibrational features of beryl
and to assign precisely the presented bands to
the vibrations. There have been many studies
using factor group analysis to calculate lines
(Adams & Gardner, 1974, [1] Hofmeister et al.,
1987, [2] Kim et al., 1995 [3]). Nevertheless,
assignment of observed bands to certain
vibrations was always one of the most
challenging tasks in vibration spectroscopy,

_______
∗

Corresponding author. Tel.: 84-4-35587061
E-mail:

32


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

moreover, the Raman shift appeared at lower
frequency in the samples with higher silicon
content and the IR band was at a higher position
in these samples. Such relations were not found
between these bands and Be concentration. We
were therefore able to confirm the assignment
of these bands to Si-O vibration.


2. Material and experimental methods
Narural beryls from Brazil (Carnaiba,
Capoeirana, Itabira, Santa Terezinha, Socoto),
Colombia (Chivor), Austria (Habachtal), Russia
(Ural), Madagascar (Mananjary), South Africa
(Transvaal),
Zambia
(Kafubu),
Nigeria
(Gwantu), China (Malipo) and synthetic ones
from Tairus, Biron (hydrothermally-grown),
Gilson, Chatham, Lennix (flux-grown) were
collected in order to cover a wide range of
chemical components. Eighty single crystals
and facetted stones were chosen for Raman
measurement and Raman spectra were obtained
from their surfaces. Then, thirty six crystals
chosen from among those already analysed by
Raman underwent chemical analysis by LAICP-MS and EMPA. From the purest eighteen
inclusion-free crystals and facetted stones, 2 mg
of powder were scraped using a diamond point
for IR measurements.
All Raman spectra were recorded at room
temperature using a Jobin Yvon (Horiba group)
LabRam HR 800 spectrometer. The system was
equipped with an Olympus BX41 optical
microscope and a Si-based CCD (chargecoupled device) detector. Spectra were excited
by Ar+ ion laser emission with 514 nm as a
green laser with a grating of 1800 grooves/mm

and a slit width of 100 µm. Due to these
parameters and the optical path length of the
spectrometer a resolution of 0.8 cm-1 resulted.
The spectra acquisition time was set at 240
seconds for all measurements. Geometrical

33

factors were strongly controlled in all Raman
measurements. One polarizer was used allowing
only the laser beam with definitive vibrational
direction (N-W) to pass through. Experiments
were then conducted with the normal
orientations of the beryl crystal (i.e orientations
of c axis) with regard to E, the electric vector.
IR spectra of beryls were recorded using a
PERKIN ELMER FT-IR Spectrometer 1725X
with 100 scan and 4 cm-1 resolution. The
samples were prepared as pellets made out of 2
mg of powdered beryl mixed with 200 mg KBr
powder to minimize the polarization effects.
Peak analysis for both IR and Raman
measurements was performed with an Originlab 7.5 professional software package. The
single and overlapping peaks were smoothed
using the Lorentz-Gauss function.
Chemical analyses were carried out by
means of LA-ICP-MS and EMPA. The use of
LA-ICP-MS served to identify Li, Be, B, Na,
Mg, Al, P, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni,
Ga, Ge, Rb, Sr, Y, Zr, Nb, Mo, Cs, Ba, La and

Ta. EMPA was used to identify the main
element Si and other elements as well in order
to have a reference matrix between LA-ICP-MS
and EMPA measurements.
Ablation was achieved with a New Wave
Research UP-213 Nd:YAG laser ablation
system, using a pulse repetition rate of 10 Hz
and 100 µm crater diameters. Analyses were
performed on an Agilent 7500ce inductively
coupled plasma - mass spectrometer in pulse
counting mode (one point per peak and 10 ms
dwelling time). Data reduction was carried out
using Glitter software. The amount of material
ablated in laser sampling was different for each
spot analysis. Consequently, the detection limits
were different for each spot and were calculated
for each individual acquisition. Detection limits
generally ranged between 0.001 and 0.5 ppm
(µg/g). 28Si was used as the internal standard.


34

L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

Analyses were calibrated against the silicate
glass reference material NIST 612 using the
values of Pearce et al. [12], and the US
Geological Survey (USGS) glass standard
BCR-2G was measured to monitor accuracy.

Microprobe analyses were achieved with a
JEOL JXA 8900RL - electron beam microprobe with wavelength dispersive analysis
technique. The chemical composition of each
sample was then corrected by PAP program.
The samples were measured by an acceleration
voltage of 20 KV and 20 nA filament current.
The detection limits differed for each element
and were affected by the overall composition of
a sample and the analytical conditions. For
most elements, the detection limit for
wavelength-dispersive (WD) spectrometers was
between 30 and 300 parts per million. The
precision depended on counting statistics,
particularly the number of X-ray counted from
the standard and sample, and the reproducibility
of the WD spectrometer mechanisms. The
minimum obtainable precision was about 0.5
percent, although it was higher for elements at
trace concentrations. Therefore, EMPA was
specially used in this study for detecting main
elements.

3. Results and Discussion
3.1. Raman band at 1076-1072 cm-1
As introduced, this band has been attributed
to the Si-O bond in the studies of Adams and
Gardner [1], Charoy et al. [6] but to the Be-O
bond in the studies of Kim et al. [3], Moroz et
al. [5] instead. According to our experimental
results, in all synthetic beryls the position of

this band was around 1067-1068 cm-1 , in
Colombian and Nigerian samples the Raman
shift was around 1068-1070 cm-1 and in
samples from Austria, Brazil, China,

Madagascar, Russia, South Africa, Zambia, the
Raman shift was around 1069-1072 cm-1. In
other words, this band shifted to higher
frequency in natural samples than in synthetic
ones (Figure 1.). Moreover, the width of this
band also varied among samples of different
provenances. The FWHM varied from 11 cm-1
to 14 cm-1 in synthetic samples, from 12 cm-1 to
15 cm-1 in samples from Nigeria and Colombia
and from 17 cm-1 to 26 cm-1 in samples from
Austria, Brazil, China, Madagascar, Russia,
South Africa and Zambia. Figure 2. showed the
plot of the peak position versus the FWHM for
beryls of different origins. Based on the FWHM
values and the Raman positions of this band, we
could separate the samples studied into two
ranges: Range I including synthetic beryls as
well as natural Nigerian and Colombian ones
were those with low FWHM and low band
position; range II including all other
investigated natural beryls.
Chemical data showed that samples of
range I contained a higher amount of silicon
than those of range II. The silicon concentration
in beryls of range I varied from 65 wt% to 66,9

wt% (from 66,1 wt% to 66,9 wt% in synthetic
samples - approximately approaching the
theoretical concentration, and from 65 wt% to
66,3 wt% in Nigerian and Colombian samples)
while silicon concentration in beryls of range II
varied from about 62,5 wt% to 65 wt%. Error!
Reference source not found. showed the
correlation between the content of silicon and
band position and FWHM for beryls of
different origins. This meant that in the samples
where the silicon content was high the band
position and FWHM were low and in the
samples where the content of silicon was low
the band position and FWHM were high. We
therefore agreed with the authors who assigned
this band to vibration of Si-O, since there was
no such correlation between beryllium
concentration and band data (Figure 4.).


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

Figure 1. Raman shift at 1067-1072 cm-1 of synthetic (solid line) and natural beryls (dot line).

Figure 2. Peak positions versus FWHMs in natural and synthetic beryls from various origins.

35


36


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

Figure 3. Correlation silicon content, band position and FWHM.

Figure 4. Correlation between beryllium content, peak position and FWHM.


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

37

In addition, the concentration of alkali ions (Na, K, Cs) was also variable among samples.

Figure 5. Alkali content versus Si content in natural and synthetic beryls from different origins.

The alkali amount of synthetic beryls varied
from 0 wt% to 0,1 wt%, and from 0,1 wt% to
0,71 wt% in natural beryls of range I, from 0,
89 wt% to 1, 87 wt% in natural beryls of range
II.
The shifting and broadening (increasing in
FWHM) of the Raman band were primarily the
results of positional disorder. Since the band
shifting and broadening were seen in low
silicon-containing samples, there were actually
other elements than silicon occupying the
silicon position. The amount of positional
disorder in each sample was the amount of lost


silicon (in comparison with the ideal silicon
amount). Other elements which could substitute
Si are Al3+, Be2+, Li+, etc. Charge compensator
could be served by alkali ions (mainly Na+, K+,
Cs+) which existed in structural channels. That
meant, the lost of silicon in beryl structure had
to be compensated by other substituting
elements (Al3+, Be2+, Li+, etc.) together with
charge compensating ions (Na+, K+, Cs+). The
correlation between Si- and alkali ion contents
elucidated this fact, since in samples where the
Si content was low, the alkali content was high
(Figure 5.).


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

38

3.2. IR band at 1071-1207 cm-1

40

S ynthe tic b e ryl

T% arb. unit

30

20


1140

10

N atural b eryl

0
400

60 0

80 0

10 0 0

12 0 0

14 0 0

-1

W a ve nu m b e r (cm )
Figure 6. IR spectra in the range 400-1400 cm-1 of beryls
(red line: natural sample from China; black line: synthetic Gilson sample).

Figure 6 showed the IR spectra in the range
400-1400 cm-1 for one alkali-free beryl (Gilson
synthesis) and for one high-alkali-containing
beryl (Chinese sample). We focused on the

band at around 1200 cm-1 which has been
assigned to the vibration of Be-O by Plyusnina
[7], Plyusnina & Surzhanskaya [8] but to the
vibration of Si-O by Manier-Glavinaz et al.
[10], Hofmeister et al. [2], Adams & Gardner
[1], Gervais & Pirou [11]. This band in fact
varied in its actual position between 1171 cm-1
and 1203 cm-1 in natural beryls (low silicon
content) and between 1200 cm-1 to 1207 cm-1 in
synthetic beryls (high silicon content). A plot of

band position versus Si content showed a trend,
that in samples with high silicon content the
band shifted toward high wave numbers
(Figure 7.). In addition, this band was shown to
be clearly more slender in synthetic samples
than in natural ones. Again, both band width
and band position were related to the
concentration of silicon and did not show any
relation to beryllium content. Therefore, the
assignment of this band to Si-O vibration was
preferred rather than to Be-O vibrations. This
observation
corresponded
with
the
interpretation of the band at 1067-1072 cm-1 in
Raman spectroscopy.



L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

Figure 7. Position of IR band at 1171-1207 cm-1 versus Si content.

Figure 8. Intensity ratio of band at 1171-1207 cm-1 and shoulder at 1140 cm-1 versus Si content.

39


40

L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

One shoulder at about 1140 cm-1 was seen
only in natural samples (with the exception of
beryls from Nigeria where the alkali content
was lower than 0,2 wt%) and was not seen in all
synthetic samples or in samples from Nigeria.
Plot of intensity ratios of band 1200 cm-1 and
shoulder 1140 cm-1 versus Si content showed a
positive linear trend, i.e. this intensity ratio was
high in samples with a high Si content (Figure 8.).
Therefore, not only band 1200 cm-1 but also
shoulder 1140 cm-1 had a relationship with the
Si content. Similarly, the plot of ratios of the
band at 1200 cm-1 and the shoulder at 1140 cm-1
versus the alkali contents showed a negative

linear trend, i.e. this intensity ratio was high in
samples with low alkali content (Figure 9. ).

Therefore, the existence of the shoulder at 1140
cm-1 in all natural samples (except Nigerian
ones) could also be related to alkali ions. The
existence of this shoulder could be explained as
follows: 1. The shoulder was generated by a
vibration X-O in which X was a divalent or
trivalent cation substituting in the Si position.
The charge compensation was served by alkali
ions (Na, K, Cs) in the channel. 2. The shoulder
was generated by M-O in which M was the
alkali ion in the channel.

Figure 9. Intensity ratio of band at 1171-1207 cm-1 and shoulder at 1140 cm-1 versus alkali content.


L.T.T. Huong, T. Häger / VNU Journal of Science, Earth Sciences 26 (2010) 32-41

4. Conclusion
In this study, based on chemical data we
have shown that the features of the Raman band
at 1067-1072 cm-1 and the IR band at 10711207 cm-1 depended on the concentration of
silicon in the sample. We therefore agreed with
the authors who assigned these bands to the
vibrations of Si-O bonding. Moreover, by using
features (FWHM, position) of these bands one
is able to separate synthetic stones which were
grown in free-alkali media from natural ones.
Raman spectroscopy as a non-destructive
method could be specially used in identification
between natural gem and synthetic beryl, since

in synthetic samples the position/FWHM of
Raman band is at 1067-1068 cm-1/ 11-14 cm-1
while these are very variable in natural ones:
1068-1072 cm-1/12-26 cm-1, respectively.

Acknowledgements
This research was financed by the Johannes
Gutenberg-University Fund for Gemstone
Research and by German Academic Exchange
Service (DAAD). Analytical facilities were
provided by the Faculty of Pharmacy,
Chemistry and Geosciences at Johannes
Gutenberg-University. The authors are grateful
for the supports.

References
[1] D.M. Adams, I.R. Gardner: Single-crystal
vibrational spectra of beryl and dioptase. Journal
of the Chemical Society - Dalton Transactions
(1974) 1502.

41

[2] A.M. Hofmeister, T.C. Hoering, D. Virgo:
Vibrational
spectroscopy
of
beryllium
alumosilicates: Heat capacity calculation from
band assignments. Physics and Chemistry of

Minerals, 14 (1987) 205.
[3] C. Kim,, M.I. Bell, D.A. McKeown: Vibrational
analysis of beryl (Be3Al2Si6O18) and its
constituent ring (Si 6O18). Physica B 205 (1995)
193.
[4] L. Nasdala, D.C. Smith, R. Kaindl, M.A.
Ziemann:
Analytical
perspectives
in
mineralogical research. EMU Notes in
Mineralogy 6 (2004) 281.
[5] I. Moroz, M. Roth, M. Boudeulle, G. Panczer:
Raman microspectroscopy and fluorescence of
emeralds from various deposits. Journal of
Raman Spectroscopy, 31 (2000) 485.
[6] B. Charoy, P. De Donato, O. Barres, C. PintoCoelho: Channel occupancy in an alkali-poor
beryl from Serra Branca (Goias, Brazil):
Spectroscopic
characterization.
American
Mineralogist 81 (1996) 395.
[7] I.I. Plyusina, and E.A. Surzhanskaya: IR
spectrum of beryl. Zhurnal Prikladnoi
Spektroskopii, 7 (1967) 917.
[8] I.I. Plyusnina: Infrared absorption spectra of
beryls. Geokhimiya 1 (1964) 31.
[9] C. Aurisicchio, O. Grubessi, P. Zecchini:
Infrared spectroscopy and crystal chemistry of
the beryl group, The Canadian Mineralogist 32

(1994) 55.
[10] V. Manier-Glavinaz, R. Couty, M. Lagache: The
removal of alkalis from beryl: Structural
adjustments. Canadian Mineralogist 27 (1989)
663.
[11] F. Gervais, and B. Piriou: Étude des spectres de
réflexion infrarouge du beryl dans les région
280-1400 cm-1. Comptes rendus de l’Académie
des Sciences Paris, B274 (1972) 252.
[12] N.J.G. Pearce, W.T. Perkins, J.A. Westgate,
M.P. Gorton, S.E. Jackson, C.R. Neal, S.P.
Chenery: A compilation of new and published
major and trace element data for NIST SRM 610
and NIST SRM 612 glass reference materials.
Geostandards Newsletter 21 (1997) 115.