X-RAY SPECTROMETRY

X-Ray Spectrom. 2006; 35: 154–158

Published online 7 April 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/xrs.886

Chemical analysis of phosphate rock using different

methods — advantages and disadvantages
Madian Jamil Safi,1∗ M. Bhagwanth Rao,1 K. Surya Prakash Rao2 and Pradip K. Govil3
1
University College of Technology, O.U., Hyderabad-500 007, A.P., India
2
Applied Geochemistry (R&T), O.U., Hyderabad-500 007, A.P., India
3
NGRI, Hyderabad-500 007, A.P., India

Received 30 May 2005; Revised 19 December 2005; Accepted 21 December 2005

Many analytical methods are considered for chemical analysis of phosphate rock because the accuracy
of analysis is very important. In the present investigations, spectrophotometric and x-ray spectrometric
methods were used to determine P2 O5 whereas SiO2 , CaO, MgO, Al2 O3 and Fe2 O3 were determined by
gravimetric and x-ray spectrometric methods. Volumetric and x-ray spectrometry methods are necessary
for fluorine determination. The investigations revealed that x-ray spectrometry is a precise and accurate
technique, which is cost effective and so better than other methods. Spectrophotometric, volumetric and
gravimetric methods are expensive, time consuming and laborious. X-ray spectrometry is suggested for
commercial companies using regular analyses to save chemicals, time and labour and to maintain quality.
Copyright  2006 John Wiley & Sons, Ltd.

INTRODUCTION using the ratio of the element peak to the Rh K˛ Compton

peak (for I, the Rh Kˇ Compton peak was used instead). The
Several methods are used for the determination of P2 O5 ,
relative standard deviation was <1.0% for the major elements
CaO, SiO2 , Fe2 O3 , Al2 O3 , MgO and F contents in phosphate
(except C, Na and Cl). The accuracy of this method was tested
rocks. However, the most reliable method used in the past
by evaluating analyses of three certified reference materials.
is the spectrophotometric determination of P2 O5 . Atomic
The direct analysis of major and minor elements in geo-
absorption spectrometry has been applied to the determina-
logical materials by the pressed pellet method without any
tion of MgO, CaO, SiO2 , Fe2 O3 and Al2 O3 1 and gravimetric
chemical procedures makes XRF spectrometry a particularly
methods for the determination of P2 O5 , CaO, SiO2 , R2 O3 environmentally friendly analytical technique.6 A detailed
(D Fe2 O3 C Al2 O3 ) and MgO. The potentiometric method chemical characterisation of phosphatised basaltic hyalo-
with a fluoride-selective electrode has long been used for the clastites assigned at Sal Island, Cape Verde archipelago, was
determination of fluorine with high accuracy. Fluorine has undertaken using synchrotron radiation x-ray fluorescence.7
also been determined by volumetric methods.1 – 3 However, Two different methods of microanalysis, energy-disper-
these analytical methods are old and outdated for critical sive spectrometry (EDS) and proton-induced x-ray emission
analysis. Major and minor oxides such as SiO2 , Al2 O3 , TiO2 , (PIXE), were used to determine the appearance of trace
CaO, MgO, KiO2 , MnO, P2 O5 and Fe2 O3 and a few selected elements (Zn, Sr and Fe) present in bone at the implantation
trace elements, Zr, Y, Sr, Rb, V, Ba, Cr, Ni, Co, Cu and Zn in site containing the ceramic. The spectra obtained by the PIXE
IWG standards AN-G, BE-N and MA-N, were determined method showed two regions for each element characterizing
relative to USGS geological reference samples and synthetic either the bone tissue or the ceramic. The PIXE method was
standards by using a PW1400 XRF analyser with a P851 apparently sufficiently sensitive for monitoring the amount
computer system in an attempt to obtain reliable values for of trace element appearing in bone-implanted material.8
the standards.4 Further, the technique is extremely useful Modern chemical laboratories use XRF spectrometry for
for analyses in mining, mineralogy and geology and in envi- the analysis of phosphate rock samples. We used the latest
ronmental work and waste material analysis. XRF is a very XRF spectrometric and also older conventional methods to
useful analytical technique in research and pharmacy.5 A compare the results.
method has been developed to determine the major, minor
and trace elements in seamount phosphorite using a modern
EXPERIMENTAL
XRF spectrometric technique. For elements having an ana-
lytical line lower in energy than the energy of the Fe K˛ line Collection and Preparation of Specimens for
absorption edge, the inter-element absorption-enhancement Analysis
effects were corrected using an influence coefficient method. The standard specimen SRM 120c was designated S1 .
For the other elements, the matrix effects were corrected One specimen of soft phosphate collected from the lower
bed in the southeastern side in the Eastern ‘B’ mines in Syria
Ł Correspondence to: Madian Jamil Safi, University College of was coned and quartered and dried. The fraction
0.015 to
Technology, O.U., Hyderabad-500 007, A.P., India. C0.037 mm was selected for wet screening and subjected to
E-mail: med [email protected] grinding to
300 mesh and this specimen was designated S2 .

Copyright  2006 John Wiley & Sons, Ltd.

 Chemical analysis of phosphate rock 155

Another specimen was collected from the tailings located measuring at the wavelength of the specimen with respect to
in the Eastern ‘B’ mines after primary concentration in the the wavelength of the standard that is used.
mines by coning and quartering. The specimen was dried A volumetric method2 for comparison with the XRF
and thoroughly mixed. The fraction
0.105 to C0.037 mm spectrometric method was applied to the determination of
was ground to
300 mesh and this specimen was designated fluorine in the phosphate rock specimens.
S3 .
Two specimens (one of them SRM 120c) were homoge-
RESULTS AND DISCUSSION
nized and ground to
300 mesh particle size and the resultant
specimen was designated S4 . One gram of the powdered The standard 120c supplied by the National Institute of
specimen (
300 mesh) was used with a backing of boric Standards and Technology (NIST) (Gaithersburg, MD, USA)
acid powder. The specimen was pressed by using a Model was analysed by XRF and other methods (P2 O5 was
TP 2d hydraulic press (Herzog Maschinenfabrik, Osnabruck, determined by spectrophotometry, CaO, MgO, R2 O3 and
Germany) (DIN EN ISO 9001 certified) to obtain pellets of SiO2 by gravimetry and fluorine by the volumetric method)
40 mm diameter. The pellets were pressed at 20 t pressure and the results are given in Table 4. Considering the absolute
and were subjected to XRF analysis. errors given by NIST, the Fe2 O3 , Al2 O3 , MgO, P2 O5 , SiO2 and
CaO contents determined by XRF analysis are similar to those
Instrumentation given by NIST. However, in other methods the MgO, P2 O5
XRF experiments were carried out with a MagiX PRO sequen- and CaO contents have absolute errors of 0.07, 0.14 and 0.32,
tial XRF spectrometer (Philips, Eindhoven, The Netherlands). respectively, while the relative errors are 0.2, 0.004 and 0.006,
A dedicated computer was used to carry out the regression respectively. Also, SiO2 was determined as insoluble residue
analysis to prepare calibration curves using the standards (IR) by gravimetric method with a similar result. Almost all
120c and 694 with reference specimens provided by the of the impurities in the insoluble residue (IR) other than SiO2
General Company of Phosphate and Mineral Resources are volatiles and were evaporated. Moreover, the reagent
(GECOPHAM) (Homs, Syria). The numbers and chemical used to precipitate SiO2 will form a complex with fluorine
compositions of the standards are given in Table 1. The and prevent the reaction between fluorine and silica. The
numbers and chemical compositions of the reference spec- gravimetric method considered insoluble residue as SiO2 ,
imens provided by GECOPHAM are given in Table 2. The which leads to inaccurate results, and it can be seen in
software in the computer was able to make dead time correc- Table 4 that the absolute and relative errors are 0.61 and
tions, background and line overlap corrections and matrix 0.11, respectively. Using the XRF spectrometry method, the
effects, giving the output directly as the concentration in F content has the absolute and the relative errors of 0.02 and
weight percentage or in ppm after converting the counts into 0.005, respectively. With the gravimetric method the R2 O3
concentration with the help of calibration curves. The con- (D Al2 O3 C Fe2 O3 ) content has absolute and relative errors
ditions under which the instrument was operated and the of 0.1 and 0.04, respectively. Using the volumetric method,
parameters for the measurement of each element are listed the F content has absolute and relative errors of 1.2 and 0.31,
in Table 3. respectively.
In gravimetric methods, the component to be determined The specimen S2 was analysed by XRF and other methods
is separated from other components of the specimen by (the same as for specimen S1 ) and the results are given in
precipitation. The weight of the precipitate was determined Table 4. Comparison of the results for specimens S1 and
after it had been filtered from solution, washed and dried.2,3 S2 showed that the gravimetric and XRF methods gave
Many processes were tried under controlled conditions, such close values of P2 O5 content. However, again the difference
as preparation of required reagent, boiling, filtration, wash- became large for the fluorine and SiO2 contents.
ing the precipitate, continuous agitation and temperature The specimen S3 was analysed by XRF and other methods
regulation. (the same as for specimens S1 and S2 ) and the results are
In this work, the P2 O5 content was determined using a given in Table 4. There is good agreement between the XRF
BCG spectrophotometer and a standard solution of P2 O5 and volumetric methods for the determination of fluorine
from monopotassium phosphate (potassium dihydrogen content. The other analytes showed significant differences
phosphate; KH2 PO4 ). In this method, P2 O5 is determined by and it can be observed that for low grades of phosphate the

Table 1. Chemical composition of the standard references

Number and identification Constituent Concentration (wt%) Constituent Concentration (wt%)

120c (Florida Phosphate Rock) Al2 O3 1.30 š 0.04 Fe2 O3 1.08 š 0.03
CaO 48.02 š 0.17 P2 O5 33.34 š 0.06
MgO 0.32 š 0.01 SiO2 5.5 š 0.2
F 3.82 š 0.02
694 (Western Phosphate Rock) Al2 O3 1.8 š 0.1 Fe2 O3 0.79 š 0.06
CaO 43.6 š 0.4 P2 O5 30.2 š 0.1
MgO 0.33 š 0.02 SiO2 11.2 š 0.4
F 3.2 š 0.1

Copyright  2006 John Wiley & Sons, Ltd. X-Ray Spectrom. 2006; 35: 154–158
156 M. J. Safi et al.

Table 2. Chemical composition of the references provided by GECOPHAM

Number and identification Constituent Concentration (wt%) Constituent Concentration (wt%)

Concentrated Syrian phosphate (Khneifiss-1SF) Al2 O3 0.2 š 0.03 Fe2 O3 0.34 š 0.03
CaO 49.3 š 0.14 P2 O5 30.8 š 0.08
MgO 0.52 š 0.01 SiO2 6.8 š 0.2
F 3.7 š 0.05
Concentrated Syrian phosphate (Eastern A-2SF) Al2 O3 0.4 š 0.1 Fe2 O3 0.28 š 0.05
CaO 46.4 š 0.3 P2 O5 28.4 š 0.1
MgO 0.63 š 0.02 SiO2 7.5 š 0.3
F 2.2 š 0.1
Concentrated Syrian phosphate (Eastern A-3SF) Al2 O3 0.3 š 0.1 Fe2 O3 0.47 š 0.03
CaO 51.9 š 0.2 P2 O5 29.7 š 0.07
MgO 0.55 š 0.02 SiO2 4.3 š 0.3
F 3.76 š 0.09
High-grade Syrian phosphate (4SF) Al2 O3 0.26 š 0.04 Fe2 O3 0.29 š 0.06
CaO 51.89 š 0.13 P2 O5 37.19 š 0.1
MgO 0.33 š 0.02 SiO2 5.38 š 0.4
F 4.03 š 0.1
Low-grade Syrian phosphate (5SF) Al2 O3 0.70 š 0.02 Fe2 O3 0.92 š 0.04
CaO 52.02 š 0.17 P2 O5 15.42 š 0.05
MgO 1.25 š 0.02 SiO2 7.34 š 0.2
F 2.97 š 0.02
Moderate-grade Syrian phosphate (6SF) Al2 O3 0.38 š 0.02 Fe2 O3 0.31 š 0.03
CaO 56.26 š 0.12 P2 O5 23.21 š 0.05
MgO 0.35 š 0.02 SiO2 4.15 š 0.2
F 3.26 š 0.08
Low-grade Syrian phosphate (7SF) Al2 O3 1.04 š 0.02 Fe2 O3 0.90 š 0.02
CaO 49.83 š 0.08 P2 O5 19.15 š 0.04
MgO 1.56 š 0.02 SiO2 11.77 š 0.1
F 3.26 š 0.05
Moderate-grade Syrian phosphate (8SF) Al2 O3 0.36 š 0.02 Fe2 O3 0.26 š 0.02
CaO 49.86 š 0.17 P2 O5 26.12 š 0.05
MgO 0.87 š 0.02 SiO2 9.44 š 0.2
F 3.70 š 0.02

Table 3. XRF analysis and the parameters for the measurement of each elementa

Offset Background
Channel 1 (Bg1)
Channel X-tal Collimator (µm) Detector kV mA Angle (° 2
) (° 2
)

Si PE 002-C 550 Flow 40 60 108.9726 2.5206

Al PE 002-C 550 Flow 40 60 144.8606 2.5584
Fe PX9 150 Scint. 30 40 57.4932 0.9596
Mn PX9 150 Scint. 40 80 62.9406 0.9760
Mg PX1 150 Flow 40 60 23.1016 2.3114
Ca PX9 150 Flow 40 70 113.1196
1.0718
Na PX1 550 Flow 40 70 27.9086 2.3812
K PX9 150 Flow 50 40 136.6522 2.0056
Ti PX9 150 Flow 40 70 86.1484 0.9296
F PX1 550 Flow 30 70 43.6202 2.1876
P Ge 111-C 550 Flow 40 70 141.0412 1.9026
a
The other conditions are same for all the elements: anode, rhodium; type, Goniometer; line, K˛; tube filter, none; measuring
time, 20 s.

Copyright  2006 John Wiley & Sons, Ltd. X-Ray Spectrom. 2006; 35: 154–158
 Chemical analysis of phosphate rock 157

Table 4. Chemical analyses of S1 , S2 , S3 , S4a and S4

Specimen Chemical analyses P2 O5 (%) CaO (%) MgO (%) Fe2 O3 (%) Al2 O3 (%) SiO2 (%) F (%)

S1 XRF method 33.29 48.08 0.32 1.09 1.29 5.59 3.78

Other methodsa 33.14 48.57 0.40 R2 O3 D Al2 O3 C Fe2 O3 D 2.55 ³ IR ³ 4.87 2.60
NIST analysis 33.34 š 0.06 48.02 š 0.17 0.32 š 0.01 1.08 š 0.03 1.30 š 0.04 5.50 š 0.02 3.82 š 0.02
S2 XRF method 25.66 52.98 0.34 0.26 0.33 7.35 3.48
Other methodsa 25.25 49.59 0.49 R2 O3 D Al2 O3 C Fe2 O3 D 0.49 ³ IR ³ 3.23 2.90
S3 XRF method 17.36 57.72 0.37 0.32 0.46 5.84 2.93
Other methodsa 19.52 51.69 0.49 R2 O3 D Al2 O3 C Fe2 O3 D 0.49 ³ IR ³ 3.23 2.90
S4a XRF method 23.37 51.17 1.73 0.29 0.39 3.48 3.50
S4 XRF method 26.63 50.23 1.26 0.59 0.81 4.49 3.61
Calculatedb 26.66 50.13 1.26 0.55 0.69 4.15 3.61
a
P2 O5 content was determined by spectrophotometry, CaO, MgO, R2 O3 and SiO2 contents by gravimetry and fluorine by the
volumetric method.
b
The contents of S4 were calculated by component balance using the mass fractions of the specimens.

gravimetric method became inaccurate and gave incorrect Table 5. Accuracy of XRF for the determined and
results for the tailings (phosphate rejected in the mines). This undetermined elements
is dolomitic phosphate with a high CaO content of about
Analyte Range of content (%) Range of variation (š)
57.72% and a high CaO/P2 O5 ratio, i.e. 3.32, as found by
XRF, but the gravimetric method gave a 51.69% CaO content SiO2 4.1–11.2 0.2–0.4
and the CaO/P2 O5 ratio of 2.65. Al2 O3 0.2–1.8 0.1–0.04
The specimen S4 was prepared by taking 2 g of a Fe2 O3 0.2–1.1 0.03–0.06
specimen (S4a ) analysed by XRF spectrometry and blending MnO 0.01–0.03 0.001–0.002
it with 1 g of SRM 120c so that the mass fractions of the MgO 0.3–1.6 0.01–0.02
specimens were 0.67 and 0.33, respectively. The specimen CaO 43.6–56.3 0.08–0.4
obtained was ground to
300 mesh particle size and Na2 O 0.8–0.5 0.001–0.02
analysed by XRF spectrometry; the chemical analyses of K2 O 0.1–0.5 0.004–0.02
the resultant specimen (S4 ) are given in Table 4. From TiO2 0.1–0.02 0.004–0.006
the component material balance and comparison with the F 2.2–4.0 0.02–0.1
analytical results for the specimen obtained (means between P 2 O5 15.0–37.0 0.04–0.1
the experimental and the theoretical results), the same values
for the F and MgO contents were found for both the
(i.e. tailings rejected in the mines) so it cannot be used for
calculated and analytical results. The absolute errors for
dolomitic phosphate where the P2 O5 content is low and that
P2 O5 , SiO2 , Al2 O3 , Fe2 O3 and CaO were 0.03, 0.34, 0.12, 0.04
of CaO is high. Even with high-grade rock (high P2 O5 con-
and 0.1 and the relative errors were 0.001, 0.08, 0.15, 0.07 and
tents) it gives inaccurate results for SiO2 . The gravimetric
0.002, respectively.
method cannot yield results for Al2 O3 and Fe2 O3 directly.
From the above results, it is clear that the XRF spectro-
However, it gives iron oxide and aluminium oxide together.
metric method gave more accurate results and can be used
For other major components, such as P2 O5 , MgO and CaO, it
for larger range of concentrations and for all major elements
can be used to determine the concentrations of unknown ele-
in phosphate rock with a wide range of standards. In the XRF
ments. The spectrophotometric method can be utilized for the
method the standards used should have a similar chemical
determination of P2 O5 only for low and high concentrations9
composition to the specimen being analysed, e.g. phosphate and is unsuitable for intermediate concentrations.
rock with MgO contents >3.00% can not be analysed using
the given standards in this work and it may give wrong
results when dolomite pebbles with higher MgO contents CONCLUSIONS
are analysed. Different standards may be required for spec- XRF spectrometry is a rapid, accurate and precise method
imens obtained from other mines. Table 5 shows the range for the elemental analysis of phosphate rock specimens and
of the analyte elements and the range of variations, and it is requires simple and minimal preparation at low cost. It
evident that XRF shows high accuracy for the major, minor has the advantage of giving separate concentrations of each
and trace elements in phosphate rock. element not related to other contents. XRF spectrometry is
In the volumetric method, several chemical processes a non-invasive method, which means that the specimens
have to be carried out for preparation of chemicals, analysed are not destroyed or changed by exposure to x-rays
calcination, digestion, boiling, distillation, agitation neutral- and they can be saved for further use as a reference or in other
ization and titration. The volumetric method is not precise types of testing. XRF covers a wide range of concentration
for the determination of fluorine content. The gravimetric using a relatively inexpensive computer for calculations but
method becomes inaccurate with low-grade phosphate rock it requires a large number of well-analysed samples for

Copyright  2006 John Wiley & Sons, Ltd. X-Ray Spectrom. 2006; 35: 154–158
158 M. J. Safi et al.

the initial determination of the coefficients and needs an 3. Furman NH. Standard Methods of Chemical Analysis: the Elements,
expert with a good knowledge of software to operate the 6th edn, vol. 1. Van Nostrand: New York, 1962.
4. Govil PK. J. Geol. Soc. India 1985; 26: 38.
spectrometer.
5. Brouwer P. Theory of XRF: ‘Getting Acquainted with Principles’, 1st
Commercial companies require XRF spectrometry in edn. PANanalytical: Almelo, 2003.
routine analyses instead of the out-of-date methods because 6. Wang X, Li G, Zhang Q, Wang Y. Geostand. Geoanal. Res. 2004; 28:
the daily checking of multiple specimens in a short time can 81.
save chemicals, time and labour and can give precise results. 7. Figueiredo MO, da Silva TP, Veiga JP, Chevallier P. J. Phys. IV
2003; 104: 399.
8. Frayssinet P, Braye F, Weber G. J. Scanning Microsc. 1997; 19:
REFERENCES 253.
9. European Fertilizer Manufacturers Association. Production of NPK
1. National Institute of Standards and Technology. Certificates of Fertilizers by the Mixed Acid Route. Booklet No. 8. European
Analysis of Standard Reference Materials 694 Western Phosphate Fertilizer Manufacturers: Brussels, 2000; 1.
Rock and 120c Florida Phosphate Rock. SRM Standard
Reference Materials Program. NIST: Gaithersburg, MD; 2003;
www.nist.gov/srm.
2. Association of Florida Phosphate Chemists. AFPC Analytical
Methods, 6th edn, vol. 9. AFPC Publications: Bartow, FL: 1980;
37.

Copyright  2006 John Wiley & Sons, Ltd. X-Ray Spectrom. 2006; 35: 154–158