Journal Archive

1-, 4-NPs and 2-NFR were purchased from Aldrich Company Inc. (Milwaukee, WI, U. S. A.). 2-NP was kindly supplied by Professor Akihisa Hirayama, Kyoto Pharmaceutical University. Deuterated 1-NP (1-NP-d₉) was purchased from CDN Isotopes Inc. (Pointe-Claire, Quebec, Canada). All other chemicals used were of analytical reagent grade.

Three on-line reducers, including an electrochemical reactor and columns packed with metals have been reported for reducing NPAHs to their corresponding amino-derivatives (Hayakawa et al., 2001, 1993; Imaizumi et al., 1990). The electrochemical reducer showed the lowest reduction efficiency and the other two showed quantitative reduction efficiency. The life time of the zinc reducer column was limited, while the life time of the Pt/Rh reducer column was not (Hayakawa et al., 2001, 1993). Considering these advantages of the Pt/Rh reducer column, we set up the HPLC-CLD system with a Pt/Rh reducer column according to our previous report (Tang et al., 2003) with several modifications (Fig. 1). The system consisted of four HPLC mobile phase pumps, a chemiluminescence reagent solution pump, a chemiluminescence detector, a system controller, an integrator, a degasser, two column ovens for the reducer column (80°C) and for the clean-up and separator columns (20°C), an auto sample injector (100 μL), a switching value, a guard column, a concentrator column, a clean-up column, a reducer column (4.0 i.d.×10 mm, packed with Pt/Rh, Shimadzu) and two separator columns (Cosmosil 5C18-MS-II, 4.6 i.d.×150 mm, Nacalai Tesqu and 5C18-AR-II, 4.6 i.d.×250 mm, Nacalai Tesque) connected in series.

Fig. 1. 
Schematic diagram of the proposed HPLC-CLD system.

The mobile phases for the clean-up column and reducer column was ethanol-acetate buffer (pH 5.5) (3:1, v/v) at a flow rate of 0.3 mL min^-1. The mobile phase for the concentrator column was the same ethanol-acetate buffer with 30 mM ascorbic acid at a flow rate of 1.8 mL min^-1. The mobile phase for the separator columns was a mixture of acetonitrile and 10 mM imidazole buffer (pH 7.6) (1:1, v/v) at a flow rate of 1.0mL min^-1. The chemiluminescence reagent solutions were acetonitrile solutions containing 0.4 mM bis(2,4,6-trichlorophenly) oxalate (TCPO) and 30 mM hydrogen peroxide, respectively, each cooled with ice-water. Both solutions were mixed (1:1, v/v) in the system, and the flow rate of the mixture was 1 mL min^-1. Other conditions were the same as in our previous reports (Tang et al., 2005, 2003).

Stock standard solutions of 1-, 2-, 4-NPs, 2-NFR and 1-NP-d₉ were independently prepared by dissolving each crystal in ethanol. The stock standard solutions were mixed, and then the mixture was diluted adequately with acetonitrile-water (3:1, v/v) to give calibration curve solutions: 0.1-100 pmol L^-1 for 2-NFR; 0.4-400 pmol L^-1 for 2-NP; 0.05-50 pmol L^-1 for 1-NP. 1-NP-d₉ standard solution was at the concentration of 0.2 pmol L^-1 in ethanol.

Airborne particulate samples were collected at Kanazawa University Wajima Air Monitoring Station at Noto Peninsula (Nisifutamata-machi, Wajima City, Ishikawa Prefecture, Japan). Airborne particulates were collected by a high volume air sampler (AH-600, Shibata Japan) with a quartz fiber filter (8×10 inch, 2500QAT-UP, Pallflex Products, Putnam, CT, U. S. A.) at a flow rate of 0.7m³ min^-1 in January and July, 2007. The filter was newly changed every week. Airborne particulates were collected from January 4 to 11, 2008, too. One-fourth of the filter (corresponding to 3.53×10³m³ air) was cut into small pieces in a flask and 100 μL of 1-NP-d₉ standard solution was added to the flask as an internal standard. Then, NPAHs were extracted ultrasonically twice with benzene/ethanol (3:1, v/v). The solution was washed successively with sodium hydroxide, sulfuric acid solution and water. After filtering the organic solution with an HLC-DISK membrane (pore size 0.45 μm, Kanto Chemical Co., Tokyo, Japan), the solution was evaporated to dryness. The residue was dissolved in 1.0 mL of ethanol, and then an aliquot (100 μL) of the solution was injected into the HPLC-CLD system.

1-NP is one of the major NPAHs exhausted directly from combustion processes of fossil fuel, and both 2-NFR and 2-NP are major NPAHs which are secondarily formed in the air (Arey et al., 1986). These facts suggest that the simultaneous determination of 1-, 2-NPs and 2-NFR is useful for determining their atmospheric behaviors. In our previous HPLC-CLD system using the reducer column packed with Pt/Rh, however, the retention time of 2-NFR was close to that of 2-NP (Tang et al., 2005). Moreover, the retention time of 4-NP was similar to the retention times of 2-NFR and 2-NP, and the concentration of 4-NP was much lower (Murahashi et al., 1999). These observations suggest that the resolution of the NPAHs should be improved, resulting in more accurate determination.

Polymeric ODS columns are known to give better resolution of PAHs than monomeric ODS columns. However, the difference of those two-type ODS columns in the separation of amino-derivatives of PAHs (APAHs) has not been examined. We compared the retentions of amino-derivatives of 1-, 2-, 4-NPs (1-, 2-, 4-aminopyrenes, 1-, 2-, 4-APs) and 2-NFR (2-aminofluoranthene, 2-AFR) on Cosmosil MS-II (monomeric ODS) and AR-II (polymeric ODS) columns by introducing them into the HPLC-CLD system as shown in Fig. 1 without the two guard columns and the clean-up column. A mixture of acetonitrile and aqueous buffer is suitable for the peroxyoxalate chemiluminescence detection of amino-derivatives of PAHs (Hayakawa et al., 2001, 1995). However, the reduction activity of Pt/Rh occurs in hot ethanol but not in acetonitrile, suggesting that the carrier solution going into the reducer column must contain ethanol but not acetonitrile. Therefore, a mixture of ethanol and acetate buffer (pH 5.5) was used as the mobile phase for the reducer column. On the other hand, amino-derivatives of PAHs were unstable in the system. Ascorbic acid was added to the effluent from the reducer column to keep the aminoderivatives stable in our previous HPLC-CLD system (Murahashi and Hayakawa, 1997). Thus, ascorbic acid was added to the mobile phase for the concentrator column in the proposed system as described in Experimental.

When NPAH standard solutions were injected into the system, NPAHs were all reduced quantitatively to their corresponding amino-derivatives in the reducer column. However, the separation of 2-AFR and 2-AP was impossible on a Cosmosil MS-II column, although the separation of 1- and 2-APs was complete. On the other hand, a Cosmosil AR-II column completely separated 2-AFR and 2-AP but did not completely separate 2- and 4-APs. By using a combination of Cosmosil MS-II and AR-II columns, the four APAHs were separately determined as amino derivatives (Table 1). The mean concentration ratio of 4-NP (0.57 fmol m^-3) was around 1/100 or 1/6 of those of 2-NFR and 2-NP, respectively, in the urban air of Kanazawa (Murahashi and Hayakawa, 1997), suggesting that 4-NP is not a good primary or secondary formation marker such as 1-, 2-NPs and 2-NFR. Accordingly, 1-, 2-NPs and 2-NFR were selected as target compounds in the following experiments and the combination of Cosmosil MS-II column (4.6 i.d.×150 mm)+Cosmosil AR-II column (4.6 i.d.×250 mm) was used for the following experiments.

Next, we examined the optimal time for switching the switching valve after loading APAHs on the concentrator column of the HPLC-CLD system as shown in Fig. 1. When the standard mixture of 1-, 2-NPs, 2-NFR and 1-NP-d₉ was injected into the system, they were eluted from the reducer column in the period from 17 to 29 minutes with the following elution order; 1-NP-d₉=1-NP<2-NFR<2-NP. From this result, the concentration time (switching time of the switching valve) was set at 17-29 minutes.

The optimum composition of chemiluminescence reagent solution is known to be different for different analytes, and the chemiluminescence intensity is known to be affected by the concentration of organic solvent in the mobile phase (Hayakawa et al., 1991). In view of the fact that the optimum acetonitrile concentration in the mobile phase was 50% for the elution and separation of 1-, 2-APs and 2-AFR, we examined the optimum concentrations of TCPO and hydrogen peroxide in the chemiluminescence reagent solution. The signal to noise (S/N) ratio of 1-AP increased with increasing TCPO concentration at concentrations above 0.02mM and peaked at 0.15 mM TCPO. Furthermore, the S/N ratios of 2-AFR and 2-AP increased with increasing TCPO concentration, peaking at 0.4 mM TCPO. The S/N ratios of 2-AFR and 1-AP increased with increasing hydrogen peroxide concentration at concentrations above 4 mM and peaked at 15 mM hydrogen peroxide. On the other hand, the S/N ratios of 2-AP increased with increasing hydrogen peroxide concentration, peaking at 30 mM hydrogen peroxide. In view of the fact that the concentration of 2-NP in airborne particulates is usually lower than the concentrations of 2-NFR and 1-NP, the concentrations of chemiluminescence reagents should be optimum to obtain the highest sensitivity of 2-AP. Thus, we used 0.4 mM TCPO and 30 mM hydrogen peroxide in the following experiments.

Under the above conditions, the detection limits (S/N=3) of 2-NFR, 2-NP and 1-NP by the proposed HPLCCLD system were 4, 10 and 1 fmol, respectively. These values were better than the values obtained by the previous system. For example, the previous detection limit of 1-NP was 10 fmol (Tang et al., 2005). The calibration curves were straight in the ranges from 10 fmol to 10 pmol (2-NFR), from 50 fmol to 50 pmol (2-NP) and from 5 fmol to 10 pmol (1-NP), respectively, with correlation coefficients over 0.999 and RSDs less than 5%.

Fig. 2 shows typical chromatograms of (A) the standard mixture of 1-, 2-NPs and 2-NFR and (B) the extracts from airborne particulates collected at Kanazawa University Wajima Air Monitoring Station on the Noto Peninsula, Japan from January 4 to 11,2008. 2-NFR, 2-NP and 1-NP were chemilumigenically determined as their corresponding amino-derivatives in chromatogram B. This result suggests that the proposed HPLCCLD system is suitable for the determination of NPAHs in air samples collected at background monitoring sites. The calculated concentrations of 1-NP, 2-NP and 2-NFR were 1.8 fmol (450 fg) m^-3, 0.65 fmol (160 fg) m^-3 and 6.0 fmol (1,500 fg) m^-3, respectively. The stable baseline of chromatogram A suggests that the proposed system can determine these NPAHs at even much lower concentration.

Fig. 2. 
Typical chromatograms of (A) standard mixture of 1-, 2-NPs, 2-NFR and 1-NP-d₉ and (B) extracts from airborne particulates collected at the Noto Peninsula. (A) Airborne particulates were collected by a high-volume air sampler at a flow rate of 1,500 L min^-1 from January 4 to 11, 2008. One-fourth of the filter was used. (B) Injected amounts of 1-, 2-NPs, 2-NFR and 1-NP-d₉ were 5, 50, 50 and 5 fmol, respectively.

1-, 2-NPs and 2-NFR in airborne particulates collected in January and July, 2007 at Noto Peninsula were determined by the proposed method. The atmospheric concentrations were not more than sub pg m^-3 levels. They were compared in Table 2. The concentrations of 1-, 2-NPs and 2-NFR were higher in winter (January) than in summer (July) by the factors of 1/3.7, 1/10.5 and 1/4.1, respectively. In both seasons, the concentrations were in decreasing order, [2-NFR]>[1-NP]>[2-NP]. It has been reported that PAHs exhausted from coal-burning systems especially in winter season in North-East China were long-range transported to Noto Peninsula over the Japan Sea (Yang et al., 2007). The concentration level of 1-NP at Noto Peninsula was much lower than that at Shenyang (annual average concentration 107±81 pg m^-3 (Hattori et al., 2007)) by the factor of 3 or 4 orders of magnitude. This is the first report that the atmospheric concentrations of 1-, 2-NPs and 2-NFR at Noto Peninsula were determined. The proposed method is useful to know the behaviors of trace levels of NPAHs at background monitoring sites such as Noto Peninsula. The long-range transport of NPAHs from Asian Continent to Japan clarified by the proposed method will be reported elsewhere.