Article pubs.acs.org/ac
Pulsed Microdischarge with Inductively Coupled Plasma Mass Spectrometry for Elemental Analysis on Solid Metal Samples Weifeng Li, Zhibin Yin, Xiaoling Cheng, Wei Hang,* Jianfeng Li, and Benli Huang Department of Chemistry and the MOE Key Lab of Spectrochemical Analysis & Instrumentation, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China ABSTRACT: Pulsed microdischarge employed as source for direct solid analysis was investigated in N2 environment at atmospheric pressure. Compared with direct current (DC) microdischarge, it exhibits advantages with respect to the ablation and emission of the sample. Comprehensive evidence, including voltage−current relationship, current density (j), and electron density (ne), suggests that pulsed microdischarge is in the arc regime while DC microdischarge belongs to glow. Capability in ablating metal samples demonstrates that pulsed microdischarge is a viable option for direct solid sampling because of the enhanced instantaneous energy. Using optical spectrometer, only common emission lines of N2 can be acquired in DC mode, whereas primary atomic and ionic lines of the sample are obtained in the case of pulsed mode. Calculations show a significant difference in N2 vibrational temperatures between DC and pulsed microdischarge. Combined with inductively coupled plasma mass spectrometry (ICPMS), pulsed microdischarge exhibits much better performances in calibration linearity and limits of detection (LOD) than those of DC discharge in direct analysis of samples of different matrices. To improve transmission efficiency, a mixture of Ar and N2 was employed as discharge gas as well as carrier gas in follow-up experiments, facilitating that LODs of most elements reached ng/g.
F
metals due to the low cost and simple equipment requirement. It has been reported that this technique is only suitable for bulk analysis because of the wandering of the discharge over the sample and unavoidable large spot diameter of several millimeters.15−17 Atmospheric-pressure discharge has attracted increasing attention due to its simplified operation without any evacuated environment. By using a needle as an electrode, it is easy to achieve a stable microdischarge at atmospheric pressure because a high electric field is formed around the needle tip.18 Experiments have been carried out to study the characteristics of microdischarge, showing its nonthermal discharge with a low gas temperature and many reactive species.18,19 Although comparison between DC and pulsed glow discharge has shown that pulsed glow discharge demonstrates a better performance than DC glow discharge at low pressure for the analysis of solid samples,20 investigations are needed to understand microdischarge processes taking place at atmospheric pressure. In the present study, features of DC and pulsed microdischarge with N2 at atmospheric pressure using needle-plane geometry are investigated. Compared with argon, nitrogen was chosen as the discharge gas because of its higher specific heat and thermal conductivity. Combined with an optical spectrometer, the pulsed microdischarge offers atomic and ionic emission lines from the sample, while the DC microdischarge gives emission lines of the
or fast and precise characterization of solid samples, direct solid analysis is preferred because of its simple sample preparation procedures. Mass spectrometry is well-known for its speed and accuracy. Many mass spectrometric techniques have been developed for the direct determination of major, minor, and trace elements in solids, such as secondary ion mass spectrometry (SIMS),1 laser ionization mass spectrometry (LIMS),2,3 glow discharge mass spectrometry (GDMS),4,5 and, more popularly, inductively coupled plasma mass spectrometry (ICPMS) coupled with solid sampling techniques.6 Recently, based on the near-field enhancement technique, the source designed for MS has been developed toward the direct analysis of solid samples with lateral and in-depth resolution in the nanometer range.7 Significantly, the integration of analytical techniques such as energy dispersive spectroscopy (EDX) and electron back scatter diffraction (EBSD) with focused ion beam/scanning electron microscopy SIMS (FIB/SEM-SIMS) is an efficient means to meet the challenge posed by complex samples.8 Considering the high-sensitivity multielemental determination via ICPMS whose source works at atmospheric pressure,9 coupled techniques for solids sampling are proposed for the simplicity and rapid sample changeover. Among them, laser ablation (LA) and arc/spark are the leading methods.10−13 LA-ICPMS is a well-established method in the analysis of both conductive and nonconductive materials. This microsampling based method can achieve excellent lateral and in-depth resolution and low LODs of ng/g with a quadruple mass analyzer.14 However, the cost of a LA device is relatively high. Application of arc/spark has been the method of choice for analysis of alloys and © XXXX American Chemical Society
Received: January 30, 2015 Accepted: April 8, 2015
A
DOI: 10.1021/acs.analchem.5b00397 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry discharge gas basically. Integrated with ICPMS, pulsed microdischarge demonstrates good linearity of calibration curves in determination of elements in different matrices. Detection limits of tens ng/g were reached with N2 as discharge gas and carrier gas. To improve the transmission efficiency, the mixture of Ar and N2 was used in the experiment subsequently, resulting in LODs of ng/g.
Table 1. Typical Operating Parameters of Microdischarge Cell and ICPMS Microdischarge DC discharge
■
EXPERIMENTAL SECTION A tungsten needle, metal sample, and quartz glass chamber (20 mm i.d., 40 mm height) constitute a discharge cell shown in Figure 1. Tungsten needle tip was fabricated via a
Voltage current
∼300 V 20 mA
Pulsed discharge repetition rate pulse width maximum output voltage maximum output current distance between needle tip and sample flow rate of N2 ICPMS
1 kHz 4 μs 3000 V 3.7 A ∼ 50 μm 50 mL/min
discharge gas RF power carrier gas flow plasma gas flow auxiliary gas flow
mixture of Ar and N2 1350 W 0 L/min 16.0 L/min 1.0 L/min
N2 1350 W 1.0 L/min 16.0 L/min 1.0 L/min
particle size of 40 μm. The powders were blended using labdancer for 10 min, then loaded into a die and pressed by a hydraulic press with a pressure of 1 × 107 Pa for 5 min, forming discs with diameter of 8 mm and thickness of 1.5 mm. The front end of needle tip was maintained at a distance of ∼50 μm from the sample surface. Data acquisition was started after a predischarge of 5 s. During the microdischarge experiments, shielding measures were taken to minimize the broadcast of electromagnetic fields to the surrounding area.
Figure 1. Schematic diagram of microdischarge cell coupled to the ICP source.
electrochemical method using tungsten wire (99.9999%) with a diameter of 250 μm (Xiamen Tungsten Ltd., China),21 serving as the anode. The sample, worked as the cathode, was mounted on a micrometer stage so that its position relative to the anode was adjustable. A pulsed power supply (SY3001, Senyuan, China) with output voltage of 0−3 kV, frequency of 1 Hz to 1 kHz, and pulse width of 1−10 μs and a DC power supply (DW-P252-20ACFO, Dongwen, China) with output voltage up to 2.5 kV and maximum current of 20 mA were utilized. A 100 kΩ ballast resistor was connected to the anode output to stabilize the discharge current in DC mode. A CCD camera was used to monitor the discharge distance and the discharge process. An optical spectrometer (FLA4000+, FLIGHT, China) was employed to collect the emission spectra. Voltage and current curves were acquired and collected via a digital storage oscilloscope (WaveSurfer422, Lecroy, USA) with a high voltage probe (PPE 6 kV, Lecroy, USA) and a current probe (TCP202, Tektronix, USA), respectively. The microdischarge cell was interfaced to the central channel of quartz torch of ICPMS (Model 4500−300, HP) via a Teflon tubing of 50 cm in length (4 mm i.d., 6 mm o.d). N2 (99.99%) was chosen as the discharge gas with a flow rate of 50 mL/min for transportation. The whole cell was airtight to guarantee the discharge proceeding in N2 environment. Typical operating parameters of microdischarge and ICPMS are listed in Table 1. A set of NIST steel standards (SRM 1761−1766), a set of copper standards containing BYG 201101, 201102, 201104, 201106, 201108, and 201110 (Nonferrous Metal Mining, China) and four sets of artificial samples with pure matrices of Ti, Co, Nb, and Pb containing Ti, Co, Nb, In, Sn, Sb, La, Ta, Pb, and Bi (New Metal Materials Technology, China) were prepared to demonstrate the performance of pulsed microdischarge. Solid standards were cut into discs with the diameter and thickness of 6 mm and 1.5 mm, respectively. Artificial samples were made from the powders which have nominal purity of 99.99% and
■
RESULTS AND DISCUSSION Apart from the ICPMS, the performance of the system is mainly influenced by discharge property and transport efficiency. Compared with DC discharge, it is necessary to employ high voltage for pulsed discharge for a steady microdischarge. The output voltage waveform shown in Figure 2a was acquired with no load. Voltage and current waveforms in Figure 2b−d are obtained under various pulse widths. Oscillations seen in both the voltage and current profiles are mainly due to the inductance in the discharge circuit. It is shown that the measured voltage was largely deviated from the output voltage, which indicates the rapid transition to arc and quick exhaustion of the energy stored in the capacity of the power supply after the ignition of the discharge. The current can reach as high as 3.7 A, and its profiles do not change significantly with the variation of pulse widths. However, the depth of the crater would increase with the rise of the incident power, resulting in the undesirable molten effect. To restrict the molten effect and obtain good signal stability, a pulse width of 4 μs and frequency of 1 kHz were chosen. Since ICP could be unstable and prone to be extinguished with addition of N2, a N2 flow rate of 50 mL/min was adopted in the subsequent experiments. Sputtering/Ablation Craters. Crater images from both discharges are shown in Figure 3 using sample SRM 1761 after discharging for 3 min. As shown in Figure 3a, an irregular mark was left on the sample surface without any sputtering, which is in good agreement with appearance of the DC discharge which wanders around the sample surface as seen in Figure 3c. For pulsed discharge, an inerratic crater with diameter of about 500 μm was formed as shown in Figure 3b, due to high instantaneous energy deposited in the discharge. A constricted microdischarge was observed for pulsed discharge as shown in Figure 3d, B
DOI: 10.1021/acs.analchem.5b00397 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 2. Voltage and current characteristics of the pulsed microdischarge with (a) pulse width of 4 μs and no load, (b) pulse width of 2 μs, (c) pulse width of 4 μs, and (d) pulse width of 6 μs.
Figure 3. Scanning electron microscope images of craters after 3 min of (a) DC microdischarge and (b) pulsed microdischarge with images of microdischarge in (c) DC mode and (d) pulsed mode.
discharge in the Ar environment at the same pulsed discharge condition is shown in Figure 4. The area of the crater is much larger, and only a slight ablation can be observed due to the relatively low specific heat and thermal conductivity of Ar. It is worth noting that spark discharge is a valid solid sampling source in Ar environment, but its electrical rod has a diameter in millimeters and peak current is normally over 20 A, which generates an extended crater and makes it a viable option only for bulk analysis.23 Emission Characteristics. In an effort to understand the differences between DC and pulsed microdischarge, emission spectra of SRM 1761 were obtained using an optical spectrometer (Figure 5). The emission spectrum of DC microdischarge reveals
indicating a significant advantage in terms of discharge volume over DC discharge. It is known that the high nonuniform electric field around the needle tip mainly accounts for remarkable coneshape of the crater with the depth of 20 μm. The ablation depth is about 0.1 nm per pulse; and the ablation rate is at ∼10−11 g with electrical energy fluence of 0.3 J/cm2 per pulse. This value is equivalent to laser energy fluence for ablation.12 In addition, as shown in Figure 3b, little molten effect was observed, indicating that sample ablation was nearly free from liquid ejection. By the use of pulsed discharge, melting can be remarkably suppressed. The pulse duration is below the time needed for the sample to reach its melting temperature; it would be unlikely to form a liquid phase.22 For comparison, the surface image of the pulsed C
DOI: 10.1021/acs.analchem.5b00397 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Table 3. Major Spectral Lines in Pulsed Microdischarge species
wavelength (nm)
N2
N2+ Fe I Fe II
Figure 4. Scanning electron microscope image of crater after 3 min of pulsed discharge in Ar environment.
electronic transition
vibrational transition
normalized intensity (%)
C 3∏u − B 3∏g 1−0 C 3∏u − B 3∏g 0−0 C 3∏u − B 3∏g 0−1 C 3∏u − B 3∏g 0−2 B 2∑u+ − X 2∑g+ 0−0 3d7(4F)4s − 3d6(3H)4s4p(3P°) 3d64s2 − 3d6(5D)4s4p(3P°) 3d6(5D)4p − 3d6(5D)4d 3d6(3D)4s − 3d6(3D)4p 3d6(3G)4s − 3d6(3F2)4p 3d6(5D)4s − 3d6(5D)4p
315.6 337.0 357.6 380.4 391.1 284.40 299.45 241.03 254.06 260.98 274.93
6.3 8.9 16.2 5.2 3.6 10.6 11.2 65.1 64.2 67.2 100
Table 4. Correlation Coefficients of the Calibration Curves and LODs for Steel and Copper Standards via Pulsed Microdischarge Coupled with ICPMS matrix element 47
Ti Ti 51 V 52 Cr 53 Cr 55 Mn 59 Co 60 Ni 62 Ni 75 As 90 Zr 91 Zr 93 Nb 96 Mo 98 Mo 100 Mo 107 Ag 109 Ag 111 Cd 113 Cd 116 Sn 118 Sn 120 Sn 121 Sb 123 Sb 125 Te 128 Te 181 Ta 206 Pb 207 Pb 208 Pb 209 Bi 48
Figure 5. Typical emission spectra acquired in two types of discharges.
Table 2. Major Spectral Lines in DC Microdischarge species N2
N2+
wavelength (nm)
electronic transition
vibrational transition
normalized intensity (%)
315.6 337.0 357.6 380.4 391.1
C 3∏u − B 3∏g C 3∏u − B 3∏g C 3∏u − B 3∏g C 3∏u − B 3∏g B 2∑u+ − X 2∑g+
1−0 0−0 0−1 0−2 0−0
25.3 100 72.3 30.1 19.4
that the dominant emissions come from the first negative bands of N2+ (B2∑u+ − X2∑g+) and the second positive bands of N2 (C3∏u − B3∏g), as listed in Table 2. The emission difference between DC and pulsed microdischarge is quite obvious. Dominant emission lines of pulsed microdischarge are in the ultraviolet region (wavelength
