Applicability of Standardless Semiquantitative Analysis of Solids by

May 1, 2009 - ... Analysis of Solids by High-Irradiance Laser Ionization Orthogonal Time-of-Fight ... State Key Laboratory of Marine Environmental Sci...
0 downloads 0 Views 922KB Size
Download PDF
Anal. Chem. 2009, 81, 4343–4348

Applicability of Standardless Semiquantitative Analysis of Solids by High-Irradiance Laser Ionization Orthogonal Time-of-Fight Mass Spectrometry Quan Yu,† Rongfu Huang,† Lingfeng Li,† Lin Lin,† Wei Hang,*,†,‡ Jian He,§ and Benli Huang† The Key Laboratory for Chemical Biology of Fujian Province and Key Laboratory of Analytical Science of Xiamen University, College of Chemistry and Chemical Engineering, State Key Laboratory of Marine Environmental Science, and Department of Mechanical and Electrical Engineering, Xiamen University, Xiamen 361005, China A compact high-irradiance laser ionization time-of-flight mass spectrometry system has been developed for the multielemental analysis of solids. Helium was introduced into the ion source as a buffer gas to cool high kinetic energy ions and suppress the interference of multicharged ions. A special pulse train repelling mode was used to achieve explicit spectra. Two quantitative methods are described for the laser ionization mass spectrometry in this paper. The first of these is the routine calibration curve quantitation, in which various matrix-matched standards are required; the second method, which is based on the uniform correlation between the signal and elemental concentration of different samples, is more convenient and covers a typical dynamic range of 6 orders. All the investigations and results indicate satisfactory performance of the newly developed instrument and its applicability for simultaneous multielemental analysis of solid samples. Currently, the direct solid sampling technique receives growing interest in geochemistry, environment, mineralogy, and many other areas, while conventional solution-based analytical techniques have been found to be defective or unusable in some cases, especially when the sample is hazardous, radioactive, or difficult to digest.1-3 Most direct solid analysis techniques require solid standards for calibration and quantification,4,5 but suitable solid standards are usually unavailable or costly for specific samples. This limits the application of solid analysis techniques. The use of lasers in solid analysis has been recognized for several advantages, such as rapid, in situ, nondestructive analysis and minimal sample preparation.6-8 There are various methods reported in some excellent references with regard to the direct * To whom correspondence should be addressed. E-mail: weihang@ xmu.edu.cn. † The Key Laboratory for Chemical Biology of Fujian Province and Key Laboratory of Analytical Science of Xiamen University. ‡ State Key Laboratory of Marine Environmental Science. § Department of Mechanical and Electrical Engineering, Xiamen University. (1) Resano, M.; Garcı´a-Ruiz, E.; Belarra, M. A.; Vanhaecke, F.; McIntosh, K. S. Trends Anal. Chem. 2007, 26, 385–395. (2) Russo, R. E.; Mao, X.; Liu, H.; Gonzalez, J.; Mao, S. S. Talanta 2002, 57, 425–437. (3) Harrison, W. W. J. Anal. At. Spectrom. 1998, 13, 1051–1056. (4) Ishida, T.; Akiyoshi, T.; Sakashita, A.; Kinoshiro, S.; Fujimoto, K.; Chino, A. Anal. Sci. 2008, 24, 563–569. (5) Schelles, W.; Van Grieken, R. J. Anal. At. Spectrom. 1997, 12, 49–52. 10.1021/ac900141z CCC: $40.75  2009 American Chemical Society Published on Web 05/01/2009

elemental analysis of solid samples by laser techniques.8-11 Since the laser-solid interaction combines both atomization and ionization, it has previously been directly coupled to time-of-flight mass spectrometry (TOFMS) to provide a viable multielemental analytical system12-14 but has limited use in qualitative analysis. TOFMS is favored for measuring ions from a laser source because of the pulsed nature of the TOF operation. High-irradiance laser ionization (LI) can generate abundant ions, but the kinetic energy of ions can be as high as several hundred electronvolts (eV),15-17 which is difficult for a regular on-axis TOF analyzer to handle.13 Resolving powers from traditional LI-TOFMS are rather poor for a high-irradiance source, while the result is barely useful for analytical purposes. Thus, we initially developed an orthogonal TOFMS (O-TOFMS) coupled with a high-vacuum LI source, but the speed of ions passing through the repelling region was so fast that only ions in a narrow mass range could be repelled into the TOF analyzer. Then, a LI-O-TOFMS was developed with an rf-only multipole ion guider added behind the low-pressure ion source.13 The resolution was enhanced substantially with the kinetic energy reduction of all ions, creating a much clearer spectrum with few multiply charged ions present, due to the recombination. Unfortunately, because of the ion guider, mass bias is unavoidable against light ions, which results in low sensitivities for light elements.18-20 (6) Russo, R. E.; Mao, X.; Borisov, O. V. Trends Anal. Chem. 1998, 17, 461– 469. (7) Omenetto, N. J. Anal. At. Spectrom. 1998, 13, 385–399. (8) Singh, J. P., Thakur, S. N., Eds. Laser-Induced Breakdown Spectroscopy; Elsevier: Oxford, U.K., 2007. (9) Lopez-Moreno, C.; Amponsah-Manager, K.; Smith, B. W.; Gornushkin, I. B.; Omenetto, N.; Palanco, S.; Laserna, J. J.; Winefordner, J. D. J. Anal. At. Spectrom. 2005, 20, 552–556. (10) Leach, A. M.; Hieftje, G. M. Anal. Chem. 2001, 73, 2959–2967. (11) Klunder, G. L.; Grant, P. M.; Andresen, B. D.; Russo, R. E. Anal. Chem. 2004, 76, 1249–1256. (12) Sysoev, A. A.; Sysoev, A. A.; Poteshin, S. S.; Pyatakhin, V. I.; Shchekina, I. V.; Trofimov, A. S. Fresenius’ J. Anal. Chem. 1998, 361, 261–266. (13) Hang, W. J. Anal. At. Spectrom. 2005, 20, 301–307. (14) Garcia, C. C.; Vadillo, J. M.; Palanco, S.; Ruiz, J.; Laserna, J. J. Spectrochim. Acta, Part B 2001, 56, 923–931. (15) Tyrrell, G. C.; Coccia, L. G.; York, T. H.; Boyd, I. W. Appl. Surf. Sci. 1996, 96-98, 227–232. (16) Ecija, P.; Rayo, M. N. S.; Martinez, R.; Sierra, B.; Redondo, C.; Basterretxea, F. J.; Castano, F. Phys. Rev. A 2008, 77, 032904/1–032904/8. (17) Apinaı`niz, J. I.; Sierra, B.; Martiı`nez, R.; Longarte, A.; Redondo, C.; Castan`o, F. J. Phys. Chem. C 2008, 112, 16556–16560. (18) Guzowski, J. P.; Hieftje, G. M. J. Anal. At. Spectrom. 2001, 16, 781–792.

Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

4343

Figure 1. Schematic diagram of LI-TOFMS.

In our newly developed instrument, collisional cooling and charge reduction are accomplished in the low-pressure source filled with inert buffer gas. The ion guider was replaced by an electrical static lens to eliminate mass bias during transportation. Many efforts were undertaken in instrumental design and parametric optimization, and as a result, ions of different masses reached quite similar sensitivities. The spectra can be used for both qualitative and quantitative analysis, especially for standardless semiquantitative analysis. Matrix effects and element fractionation are considered to be the major hindrances when a laser technique is used for quantitative analysis.7,21-23 The fundamental process of laser ablation and ionization in a low-pressure source with ambient gas has not been completely investigated, owing to the complex interactions among laser beam, neutrals, ions, electrons, and buffer gas. Some efforts based on the investigation of ion activity in our instrument have been carried out to reduce the fractionation problem, including the adoption of orthogonal ion extraction, static lens, and other designs, which will be later described in detail. The main aim of our research is to explore the potential of using LI-O-TOFMS in the quantitative analysis of various kinds of solids. The results obtained reveal that this technique is barely affected by sample matrix and has a wide dynamic range, from 0.0001% to almost 100%. This system has successfully performed a direct, rapid, and sensitive multielemental composition measurement for solid samples such as alloys and geological samples. EXPERIMENTAL SECTION A small LI-O-TOFMS has been built in-house, and the schematic diagram of this system is shown in Figure 1. A Nd:YAG laser (NL303G, EKSPLA) was used in all experiments. The operating laser wavelength was set at third harmonic (355 nm) (19) He, J.; Yu, Q.; Li, L.; Hang, W.; Huang, B. Rapid Commun. Mass Spectrom. 2008, 22, 3327–3333. (20) Tong, Q.; Yu, Q.; Jin, X.; He, J.; Hang, W.; Huang, B. J. Anal. At. Spectrom. 2009, 24, 228–231. (21) Gornushkin, I. B.; Smith, B. W.; Potts, G. E.; Omenetto, N.; Winefordner, J. D. Anal. Chem. 1999, 71, 5447–5449. (22) Kondrashev, S.; Kanesue, T.; Okamura, M.; Sakakibara, K. J. Appl. Phys. 2006, 100, 103301/1–103301/8. (23) Longerich, H. P.; Gu ¨nther, D.; Jackson, S. E. Fresenius’ J. Anal. Chem. 1996, 355, 538–542.

4344

Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

for the consideration of less fractionation in ablation.24,25 The laser beam was focused by a focusing objective (LMU-3X, OFR division of Thorlabs Inc.) to produce an ablation spot of 20 µm in diameter. The focal spot was positioned approximately 8 mm from the sampling nozzle in the X direction and 4 mm in the Y direction. Two M-661 miniature translation stages (Physik Instrumente Co.), controlled by a C-170 RedStone piezomotor driver, were stacked to act as a two-dimensional manipulator mounted at the tip of an in-house-built direct insertion probe (DIP). The stage was kept moving in the experiment to provide a fresh sample surface since excessive laser shots on the same spot result in variational ion production owing to cavity effects.26 Two pure samples, tungsten and copper, were initially used for instrument optimization. A set of steel standards were applied for quantitative analysis research, including a NIST standard (SRM 1262b) and five Chinese national standards (GBW01396, 01398, and 01399 from China Iron and Steel Research Institute Group, GSBH40115-2 and 40115-3 from National Standard Materials Research Center). These samples were all cut into disks of about 6 mm in diameter and 1.5 mm in thickness. In addition, four other standards with different sample matrixes were also used in our investigation, including copper (SRM 1112), zinc (SRM 629), and silicon dioxide (GBW (E) 070041 and 070043, Chinese Academy of Sciences). Among them, the alloy samples were treated as the steel standards, and the two ore powder samples were pressed into disks of the same shape as the steel sample under a pressure of 3 × 107 Pa. Generally, soil samples had a less homogeneous surface owing to the discrepancy in powder sizes. All sample disks were adhered onto the manipulator by carbon tape. A needle valve was used to control the pressure of the source chamber, and ultrahigh purity helium (99.999%) was injected as a buffer gas. The instrument consisted of three vacuum stages: ion source stage, transportation stage, and TOFMS stage. A nozzle with an orifice of 1 mm in diameter was used to sample ions from the source. Behind the nozzle was the transmission system that contained three compact cylinder lenses (L1, L2, L3) and a set of Einzel lenses, separated by an end plate with a center aperture (1 mm in diameter). The short transmission system (15 cm from nozzle to repelling plate) was carefully designed to avoid fractionation. The three lens elements were used to focus the ions into the TOFMS stage through the aperture, and the Einzel lens guided the ions into the repelling region. An in-house-built pulse train generator triggered the repelling pulses in TOFMS. The pulse train consisted of a certain number of pulses with a fixed frequency of 33 kHz and a duration of 2.5 µs and was initiated by the laser synchron pulse. The TOFMS with angular reflection was built in-house. The distance from the repelling plate to the backplate of the reflectron was only 50 cm, and the resolving power was typically 3000. Table 1 summarizes the detailed operating conditions of the system. Data acquisition was achieved in both time-to-digital conversion (TDC) mode and analog-to-digital conversion (ADC) mode, using a time-to-digital converter (TDC×4, Ionwerks) and a digital storage (24) Alexander, M. L.; Smith, M. R.; Hartman, J. S.; Mendoza, A.; Koppenaal, D. W. Appl. Surf. Sci. 1998, 127-129, 255–261. (25) Russo, R. E.; Mao, X. L.; Borisov, O. V.; Liu, H. J. Anal. At. Spectrom. 2000, 15, 1115–1120. (26) Jeong, S. H.; Greif, R.; Russo, R. E. J. Appl. Phys. 1996, 80, 1996.

Table 1. Typical Operating Parameters of the LI-TOFMS Source ambient gas source chamber pressure

helium 800 Pa

Laser wavelength duration repetition rate energy spot diameter irradiance

355 nm 4.4 ns 10 Hz 2 mJ 20 µm 1.4 × 1011 W/cm2 Transmission System +18 V 0V -150 V -10 V -20 V 0V -25 V 10-2 Pa

nozzle L1 L2 L3 end plate Einzel lensssides Einzel lenssmiddle pressure

TOFMS repelling pulse frequency repelling pulse magnitude acceleration potential steering plate potential pressure

33 kHz 450 V -2400 V -2315 V 10-4 Pa

oscilloscope (Lecroy, 42Xs), respectively. A fast-timing preamplifier (VT120, Ortec) and a constant fractional discriminator (model 6904, Phillips Scientific) were used only in TDC mode. Each TDC spectrum represents an accumulation of 3 × 104 scans unless specially indicated, which requires about 3 min to complete the acquisition. Correspondingly, the ADC spectrum is the sum of 50 single data measurements from the oscilloscope, which are acquired in 5 s by in-house-written software with LabView. A proper internal standard element was selected for data interconversion between these two acquisition modes. ADC was used to acquire the signals of high-concentration elements, whereas TDC was used for small signals. In quantitative analysis, the concentration information of most elements was provided by the TDC, except for some high-concentration elements. With the current setup, the instrument has the limit of detection (LOD) of sub-parts-per-million for most elements. RESULTS AND DISCUSSION The ion trajectory in a static field essentially depends on the ion kinetic energy. In a high-irradiance laser source, the diversity of the ion initial kinetic energy will result in quantification errors because of variant transmission efficiency. To solve this problem, orthogonal ion extraction geometry is introduced into the source. SIMION simulation results demonstrate that only those sufficiently cooled ions (below 1 eV in the Y direction) can be transported through the aperture of the end plate, whereas most of the energetic ions are difficult to swerve due to the orthogonal geometry. In this way, a possible extraction process was suggested

in which ions are cooled close to the thermal motion state and then diffuse into the nozzle. The kinetic energies of the sampled ions are so small that they can be ignored in comparison with the electric potential energy that is determined by the voltage applied on the nozzle (18 V). Hence, all the sampled ions have approximately the same energy. The utilization of a static lens in ion transmission optics ensures that ions with similar kinetic energy will undergo little fractionation during transport. This is supported by the uniform ion kinetic energy distribution in the X direction at the repelling region, which is measured by scanning the steering plate potential. The method of measuring the kinetic energy distribution via steering plate potential was described in ref 27. As mentioned above, ions generated through LI have kinetic energies of up to several hundred electronvolts, while a TOFMS can only handle ions with kinetic energies of less than 20 eV, even in reflectron mode.28 Buffer gas plays an important role in reducing the kinetic energy of ions. If the pressure is too low, ions have deficient collisional cooling and will not be sampled, whereas excessively high pressure may enhance the resistance and ion neutralization, resulting in signal reduction. Under the current experimental conditions, the pressure is chosen at 800 Pa after optimization, where the mean free path of the Fe ion is 0.03 mm. In fact, the collisions between the ions and buffer gas mainly take place after the expansion, and the plume length is estimated to be 3 mm according to previous research by Gonzalo et al.29 Thus, the ion-helium collision cooling region is about 7 mm in length, which means that all ions experience more than 200 collisions before being sampled. It is estimated that a Fe ion, for example, loses 12% of its kinetic energy after each collision with helium, according to the elastic collision model. Thus, most of the ions can be sufficiently cooled and sampled. The plume-nozzle orthogonal geometry in the source can facilitate the laser beam introduction and avoid energetic ions sampled into the nozzle. However, the orthogonal extraction has less sampling efficiency compared to the on-axis geometry routinely used. To enhance sensitivity, more intensive laser power density was applied in comparison with that used in previous work.13 Though the generation of multicharged ions increases with an increase in the laser energy,17,22,30,31 most of the highly charged ions are charge-reduced by frequent recombination, which is illustrated below. During LI, the amount of ions generated is strongly dependent on both the laser intensity and the surrounding conditions. After absorption of a high-irradiance laser, the target is evaporated and ionized creating a dense plasma plume above the material surface. Even in high vacuum, most laser ablation particles undergo extensive collisions and recombination in the freely expanding plasma. Electrons are preferentially attracted to multiply charged ions that have higher positive valencies than singly charged ions. (27) Hang, W.; Yan, X.; Wayne, D. M.; Olivares, J. A.; Harrison, W. W.; Majidi, V. Anal. Chem. 1999, 71, 3231–3237. (28) TOF Fundamentals. TOF Tutorial by R. M. Jordan Co. http://www. rmjordan.com/index.html. (29) Gonzalo, J.; Go´mez San Roma´n, R.; Perrie`re, J.; Afonso, C. N.; Pe´rez Casero, R. Appl. Phys. A: Mater. Sci. Process. 1998, 66, 487–491. (30) Gornushkin, I. B.; Kazakov, A. Y.; Omenetto, N.; Smith, B. W.; Winefordner, J. D. Spectrochim. Acta, Part B 2005, 60, 215–230. (31) Namba, S.; Nozu, R.; Takiyama, K.; Oda, T. J. Appl. Phys. 2006, 99, 073302/ 1–073302/8.

Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

4345

Figure 2. Transient ion profile after a single pulse from tungsten samples. The delay time axis is in logarithmic scale.

With the presence of ambient gas, recombinations take place through frequent collisional processes (mainly three-body, possibly radiative two-body or charge exchange collisions), which results in the preferential charge reduction of highly charged ions.31-34 Figure 2 shows the transient evolutions of ions from pure tungsten target at different delay times between the laser pulse and the repelling pulse. Copper transient profile is similar to the tungsten profile, except that no copper oxide peak can be observed in the spectra. The integrated peak area demonstrates that doubly charged ions constitute only a small percentage (e.g., 1.5% for Cu2+, 2% for W2+, and 4% for WO+) with respect to the total ions generated per pulse. The ratio of highly charged ions in our experiment is significantly less than that found in other investigations in high vacuum. For example, Garcia et al.14 observed that the Zn2+ to Zn+ ratio was about 50% in LI-TOFMS at a laser irradiance of 109 W/cm2; Wang et al.35 also showed 30% Y2+ to Y+ in the spectrum at 108 W/cm2 irradiance. As shown in Figure 2, the doubly charged ions show up in the front of the ion package. This is due to the front location of the plume, where highly charged ions are generated. Since plasma shielding occurs in nanosecond laser ablation, those ions located in the sheath of the plasma plume absorb the largest laser energy, where the doubly charged ions are most probably generated.36 The time profiles of singly and doubly charged ions are evidently different. Singly charged ions exist for approximately 1 ms, while doubly charged ions last only tens of microseconds. The front location and the preferential charged reduction result in the narrow profiles of the doubly charged ions. After the free expansion, the active ions and electrons undergo collisions with the abundant helium atoms. Most of electrons are recombined or escape, whereas ions are slowed down and cooled after collisions, then diffuse into the nozzle. The discrepancy between the singly and doubly charged ions in the source could result in the different distribution of sampling time. This distribution is further magnified during ion transmission.37,38 As a consequence, (32) Levashov, V. E.; Mednikov, K. N.; Pirozhkov, A. S.; Ragozin, E. N. Radiat. Phys. Chem. 2006, 75, 1819–1823. (33) Beigman, I. L.; Levashov, V. E.; Mednikov, K. N.; Pirozhkov, A. S.; Ragozin, E. N.; Tolstikhina, I. Y. Quantum Electron. 2007, 37, 1060–1064. (34) Gupta, G. P.; Sinha, B. K. Phys. Rev. E 1997, 56, 2104–2111. (35) Wang, X.; Amoruso, S.; Armenante, M.; Boselli, A.; Bruzzese, R.; Spinelli, N.; Velotta, R. Opt. Lasers Eng. 2003, 39, 179–190. (36) Harilal, S. S.; O’Shay, B.; Tao, Y.; Tillack, M. S. J. Appl. Phys. 2006, 99, 083303/1–083303/10. (37) Wilcox, B. E.; Hendrickson, C. L.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2002, 13, 1304–1312. (38) Taban, I. M.; McDonnell, L. A.; Rompp, A.; Cerjak, I.; Heeren, R. M. A. Int. J. Mass Spectrom. 2005, 244, 135–143.

4346

Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

Figure 3. Schematic diagram of repelling pulse in TOFMS detection. Different types of ions are distinguished in the ion package per laser pulse: (1) multiply charged ions, (2) singly charged ions, and (3) oxidation ions.

singly charged ions spread in about 1 ms when passing the repelling region. This stretched ion package is treated as a quasicontinuous ion beam and can be sampled 30 times by the TOF analyzer operated at a repelling frequency of 33 kHz. In current experiments, the appearance of oxidation ions (e.g., WO+ in Figure 2) are further deferred because the ion molecule reactions occur several hundred microseconds after the laser plasma.39 In the experiment, the Nd:YAG laser was operated at the maximum repetition rate of 10 Hz, leaving 100 ms for ion detection per laser pulse. The temporal ion distribution indicates that only the first millisecond after the laser pulse is useful for sampling ion signals in the repelling region. The existence of multiply charged and polyatomic ions could result in spectral interference; hence, a special repelling scheme is used in current TOFMS detection. In Figure 3, the repelling pulse for ion extraction works all the time in routine continuous mode, whereas in the novel pulse train mode, the extraction is performed only in a partial time period. The pulse number per train and the delay time between laser pulse and pulse train can be adjusted for the selective detection of singly charged ions. This strategy will enhance the signal-to-noise ratio and remove interference. Figure 4a shows the mass spectrum acquired by continuous pulses, indicating severe interference of multiply charged and polyatomic ions. In contrast, the interference is effectively suppressed in pulse train repelling mode (Figure 4b), in which a selective portion of the ion package was extracted with delay times from 100 to 900 µs. The comparison of signal intensity in Figure 4, parts a and b, demonstrated less than 5% loss of singly charged elements, which was attributed to previous investigations of the ion temporal profile. (39) Cosic, B.; Ermoline, A.; Fontijn, A.; Marshall, P. Proc. Combust. Inst. 2007, 31, 349–356.

Figure 5. Calibration curves for representative elements in steel standards.

Table 2 element 27

Al Ti V 52 Cr 55 Mn 59 Co 60 Ni 48

51

Figure 4. Comparison of mass spectra acquired in two different repelling modes: (a) continuous pulse; (b) pulse train. GBW01398 was used, and the acquisition time was 100 s containing 1000 laser pulses. The inset spectra present a mass range of less than 30.

With the elimination of most interference in the pulse train repelling, quantitative analysis of elemental composition was processed using a routine calibration method. The calibration curves were obtained using steel standards with different element compositions. Figure 5 shows some calibration plots, while all correlation coefficients (R2) and relative standard deviations (RSDs) are listed in Table 2. In practice, the RSDs were quite consistent across the different concentration ranges, so the mean values of the RSDs are shown. In addition to quantification, the results demonstrated good stability of this instrument. As seen in Figure 5 and Table 2, quantitative analysis can be performed for matrix-matched samples in LI-TOFMS. The correlation between signal magnitudes and elemental compositions is linear. It is well-known that ions generated through LI are dependent on the matrix of solids.40 Usually, a calibration curve is required for the quantization of each element. To plot calibration curves, solid standards are required and need to meet matrix-matched, specific elements contained, and element concentrations in a suitable range, which are usually unavailable or costly. Therefore, an investigation was carried out to prove the ability of LI-TOFMS in semiquantitative analysis with no standard requirement. Figure 6 shows the respondence of signal and concentration for all elements for six steel standards with only one isotope (40) Qaisar, M. S.; Pert, G. J. J. Appl. Phys. 2003, 94, 1468–1477.

R2 0.995 0.983 0.992 0.996 0.912 0.943 0.977

RSD (%) 8.4 8.9 12.0 7.2 7.7 3.4 5.6

element 63

Cu Nb Mo 184 W 208 Pb 209 Bi 93

98

R2

RSD (%)

0.999 0.992 0.999 0.996 0.968 0.920

7.1 6.9 11.3 8.9 12.6 10.0

representing each element. The fit line was calculated by the leastsquares method with ORIGIN 7.5. The composition unit used in the calibration was molar concentration (10-6 mol/g), which essentially corresponds to the atom number of the specific element ablated per pulse. For intensive laser irradiance, as in current studies, the ionization efficiency of different elements was supposed to be uniform.41,42 The correlation in Figure 6 demonstrates the detection unity for different elements and also indicates that our compact design of the whole system effectively reduces the elemental fractionation during transportation. According to the plot, an LOD value as low as 10-8 mol/g can be

Figure 6. Correlation plot of element composition and signal intensity from six steel standards: (0) SRM1262b, (4) GSBH40115-2, (]) GSBH40115-3, (b) GBW01396, (f) GBW01398, and ([) GBW01399. All metal elements measured are presented in the plot, covering a concentration range of approximately 6 orders. Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

4347

Table 3. Results of Isotopic Abundance Ratio (%) for GBW01396 isotope 54

Fe 56 Fe 57 Fe 92 Mo 94 Mo 95 Mo 96 Mo 97 Mo 98 Mo 100 Mo

composition %

natural

measured

5.09 83.56 1.95 0.15 0.10 0.17 0.18 0.11 0.27 0.11

5.8 91.8 2.1 14.8 9.3 15.9 16.7 9.6 24.1 9.6

5.7 92.0 2.3 14.2 8.7 15.3 17.1 8.4 26.9 9.3

achieved, which gives the practical detection limit of this instrument in the form of molar concentration. Moreover, the LOD value is applied to most elements owing to the fact that the signal is mainly determined by the compositions, not the elements. The commonly used LOD, in the unit of gram/gram relating to the atomic masses of the elements, is in the order of sub-parts-per-million in our experiment. In addition, a general approach for semiquantitative analysis is provided based on the linear fit curve. Isotopic compositions can be directly calculated from their signal magnitudes in Figure 6, such that isotopic analysis can be carried out. As shown in Table 3, isotopic abundance ratios are in reasonable agreement with their natural values, suggesting the feasibility of this quantification method. Since the LI technique suffers a matrix effect that limits its application in quantitative analysis, the evaluation of this effect is carried out with several non-matrix-matched samples. According to the agreement between experimental and certified composition values in Figure 7, this technique appears to be independent of matrix composition. The Lin’s concordance correlation coefficient43 is used to evaluate this measurement, which gives the Fc value of 0.986. Hence, the correlation of signal intensity and element composition can be used directly to perform semiquantitative analysis without matrix-matched standards. Usually, matrix effects can be observed at the laser irradiance of 108-1010 W/cm2.7,22,40 In our experiment, the laser irradiance applied (1.4 × 1011 W/cm2) is higher than that routinely used for LI.12-14 High irradiance produces adequately high-temperature

isotope 180

W 182 W 183 W 184 W 186 W 204 Pb 206 Pb 207 Pb 208 Pb

composition %

natural

measured

0.00263 0.54 0.29 0.64 0.60 0.00003 0.00050 0.00046 0.00110

0.13 26.3 14.3 30.7 28.6 1.4 24.1 22.1 52.4

0.10 26.8 12.9 33.4 27.8 2.0 23.4 18.3 56.2

plasma to degrade the influence of the matrix effect, as described previously.41 The use of buffer gas in the source and orthogonal instrument geometry make it possible for TOF to work properly with the energetic ions generated in high laser irradiance, which approaches the ideal LI condition. The aim of this research is to prove the ability of high-irradiance LI-TOFMS in semiquantitative multielemental analysis. All results indicate satisfactory performance of the instrument developed. The results in Figures 6 and 7 have confirmed the appropriate quantitation method for solid sample analysis. The linear calibration curve can be used to directly evaluate the elemental concentration from the signal intensity obtained in the spectra, independent of the element species and sample matrix. This technique presents a direct semiquantitative analysis method, covering a wide dynamic range of 6 orders of magnitude (in Figure 7). CONCLUSIONS The potential of high-irradiance laser ablation ionization TOFMS has been investigated in solid semiquantitative multielemental analysis. A low-pressure LI source can effectively reduce the amount of multiply charged ions through collision and recombination; buffer gas also helps to cool the energetic ions in the source, which facilitates the MS sampling. The interference of multicharged and cluster ions is further suppressed in pulse train repelling mode. High laser irradiance leads to sufficient atomization, unity ionization, and suppresses the matrix effect. Direct semiquantitative measurements can be performed according to the correlation between signal intensity and element composition, covering a wide dynamic range. With these achievements, this technique can be used for qualitative and semiquantitative solid analysis with no standard requirement. ACKNOWLEDGMENT Financial support from the National 863 program, Natural Science Foundation of China, and Fujian Province Department of Science & Technology are highly acknowledged. Received for review January 21, 2009. Accepted April 15, 2009. AC900141Z

Figure 7. Correlation plot of experimental and certified composition from five standards with different matrixes: (O) SRM1262b, (0) GBW070041, ([) GBW070043, (9) SRM 1112, and (f) SRM 629.

4348

Analytical Chemistry, Vol. 81, No. 11, June 1, 2009

(41) Vertes, A., Gijbels, R., Adams, F., Eds. Laser Ionization Mass Analysis; Wiley Inter-science: New York, 1993. (42) Matus, L.; Seufert, M.; Jochum, K. Int. J. Mass Spectrom. Ion Processes 1988, 84, 101–111. (43) Lin, L. I. K. Biometrics 1989, 45, 255–268.