1 Introduction

Nowadays, more and more people realized that environmental pollution, especially the distribution and diffusion of heavy metal pollution in the soil and water, brings a high risk to the local environment and citizens. Therefore, the development of methods for measuring heavy metals are appreciated. The conventional methods, including chemical analysis, spectrophometic method, atomic absorption spectrophotometry, and other, are very time-consuming and expensive [1], and the samples need to be collected and prepared before analyzing. To tackle these problems, PGNAA can be considered as a suitable technique for an in situ, rapid and continuous survey of such heavy metal pollutants [2].

PGNAA is a powerful nuclear analytical technique for the determination of elements. It has been applied in many fields, including on-line and in situ measurements, and environmental and non-destructive analysis [3-8]. The method is mainly based on some of the most basic nuclear reactions, thermal neutron capture and neutron inelastic scattering reactions. The number of prompt gamma-rays, which are emitted from the sample during the reactions, is proportional to the amount of elements.

To discuss the possibility of heavy metal detection in aqueous solutions using the PGNAA technique, some works have been studied by some organizations. A detection facility containing a 0.125 m³ volume of water, ²⁴¹Am-Be neutron source, and high purity germanium gamma-ray detector was designed by Khelifi et al., and some heavy metals, such as Mn, Cu, Ni, and Cr, were studied by the setup [9]. Naqvi et al. used a moderator to get the thermal neutrons from the neutron generator and then detected the samples containing Cd and Hg [10, 11]. In our previous work, a PGNAA setup for element detection of aqueous samples has been designed [12].

In this paper, heavy metals, including Mn, Cu, Ni, Cr and Zn, were detected with the PGNAA setup, which was designed in our earlier work. The linearity response between the characteristic peak counts and concentrations of heavy metals were studied. The minimum detectable concentration of each element was also discussed.

2 Experimental

2.1 PGNAA setup

A schematic diagram of the setup is shown as Fig. 1. The setup consists of a 300 mCi ²⁴¹Am-Be neutron source and a cylindrical 4 × 4 inch (diameter × height) BGO detector (Saint-Gobain, the energy resolution is 9.7% for 662 keV gamma-rays from the ¹³⁷Cs source). The neutron source is placed at the center of the setup. The BGO detector is located at the external surface of the polymethyl methacrylate tank and parallel to the vertical centerline. The detector is connected to an ORTEC model digibase multi-channel analyzer. The radius and height of the tank were set to 16 cm and 40 cm according to Monte Carlo optimization, respectively [12]. The aqueous samples, which play a very important role in moderating neutrons, are in the cavity separated from the source by a tube. The energy calibration of the setup is performed using the characteristic gamma-rays emitted from the ⁶⁰Co, ¹³⁷Cs, ²⁴¹Am-Be neutron source and prompt gamma-rays by hydrogen neutron capture.

Fig. 1.Full-screen View

Schematic diagram of PGNAA setup.

3 Results and Discussion

Figures 2(a)–2(e) show the gamma-ray spectra of the BGO detector from the aqueous samples contained in heavy metals. For each element, there were five samples containing different concentrations. The deionized water was considered a background. It can be seen that the spectra of prompt gamma-rays of the heavy metals contained different concentrations superimposed upon each other, along with the background spectrum. The characteristic gamma-rays of Mn, Ni and Cr can be seen directly, while Cu and Zn are not so clear due to the small absorption cross-section in the spectra. Next, the background spectrum was subtracted from the samples spectra. And the peaks of difference spectra were found and integrated to generate measured counts of prompt gamma-ray peaks of heavy metals [11].

Fig. 2.Full-screen View

The prompt gamma-ray spectra of heavy metals (a:Mn b:Cu c:Ni d:Cr e:Zn)

The measured counts of the chosen gamma-rays from Mn, Cu, Ni, Cr and Zn plotted against their respective concentrations are shown in Figs. 3(a)–3(e). It can be seen from the figures that the count varies linearly with the amount of each element. All coefficients of correlation of the linear fit are greater than 0.999. The absorption cross-sections of these heavy metals are small, so that the effect of neutron self-shielding can be ignored. Hence, the results show a good linear relationship between the counts of prompt peak with their concentrations. In our previous work, Cl, B, Hg and Cd were studied with the setup [12, 15]. However, the coefficients of correlation of the linear fit of these elements are poorer than that of the heavy metals due to their large neutron absorption cross-section. Neutron self-shielding has a significant influence on the elements’ measurement. Some methods have been used to correct the effect of the neutron self-shielding [14].

Fig. 3.Full-screen View

The gamma-ray counts vs. concentrations of heavy metals (a: Mn; b: Cu; c: Ni; d: Cr; e: Zn)

In order to describe the detection limits of the elements measured by this setup, the minimum detectable concentration (MDC) of the setup for these elements are calculated. The MDC and the errors are calculated by the following equations [16]:

where C is the concentration of the element of interest; P and B are the net count and background count of the characteristic peak, respectively; p and b are the net count and background count rate of the characteristic peak, respectively; t is the measured time.

The MDC of the heavy metals at the chosen characteristic energies are determined by the PGNAA setup with a time of 3600 s. The results are listed in Table 2. It can be seen that only cadmium could be obtained with the detection limit in the order of ppm, while mercury can be determined in the order of 100 ppm. The MDC of Mn, Cu, Ni and Cr are all in the range of 200 ppm–400 ppm. However, the detection limit of Zn is in the order of mg/ml due to its small absorption cross-section. To improve the detection limits of heavy metals, using a more intense neutron source or taking a longer time measurement can be considered. Both the improvements can make this facility a more powerful tool in analyzing heavy metal pollutants.

References [4] and [9] reported some minimum detection limits for heavy metals. The results are listed in Table 3. Compared to these works, the setup we designed performs well in heavy metal detection.

4 Conclusion

A PGNAA setup with a ²⁴¹Am-Be neutron source and BGO detector has been used to detect heavy metals in water. The results show that the facility can be used for the determination of some heavy metal pollutants. The detection limits for heavy metals, which are commonly found in polluted water, were evaluated and predicted to investigate the potential application of the probe for in-situ measurements. Higher detection sensitivity can be obtained by increasing neutron flux or taking a longer time measurement.