Anal. Chem. 2002, 74, 261-269
Development of a Precolumn Derivatization Method for the Determination of Free Amines in Wastewater by High-Performance Liquid Chromatography via Fluorescent Detection with 9-(2-Hydroxyethyl)acridone Jinmao You,†,‡ Weibing Zhang,† Qinghe Zhang,† Lin Zhang,‡ Chao Yan,† and Yukui Zhang*,†
Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 116011, China, and Department of Chemistry, Qufu Normal University, Qufu Shandong, 273165, China
A simple, sensitive, and mild method for the determination of amino compounds based on a condensation reaction with fluorescence detection has been developed. 9-(2-Hydroxyethyl)acridone reacts with coupling agent N,N′-carbonyldiimidazole at ambient temperature to form activated amide intermediate 9-(2-acridone)oxyethylcarbonylimidazole (AOCD). The amide intermediate (AOCD) preferably reacts with amino compounds under mild reactions in the presence of 4-(dimethylamino)pyridine (base catalyst) in acetonitrile to give the corresponding sensitively fluorescent derivatives with an excitation maximum at λex 404 nm and an emission maximum at λem 440 nm. The labeled derivatives exhibit high stability under reversed-phase conditions. The fluorescence intensities of derivatives in various solvents or at different temperatures were investigated. The method, in conjunction with a gradient elution, offers a baseline resolution of the common amine derivatives on a reversed-phase C18 column. The LC separation for the derivatized amines shows good reproducibility with acetonitrile-water including 2.5% DMF as mobile phase. The relative standard deviations (n ) 6) for each amine derivative are C10), the stock solutions were prepared by dissolving the amine in DMF and diluting with acetonitrile to a concentration of 50 ng/mL. A standard solution of the coupling agent (10 mg/mL) was prepared by dissolving 0.1 g of N,N′carbonyldiimidazole in 10 mL of acetonitrile. Standard solution of the basic catalyst (20 mg/mL) was prepared by dissolving 0.2
Table 1. Chromatographic Gradient Conditions For the Separation of Amine Derivatives (Method 1)a time (min) A 30% ACN B 70% ACN C 100% ACN 0 20 25 40
100 0 0 0
0 100 0 0
0 0 100 100
For the Separation of Amine Derivatives (Method 2) time (min) A 40% ACN B 100% ACN 0 4 20 40 a
100 100 0 0
0 0 100 100
A, B, and C contained 2.5% DMF, respectively.
g of 4-(dimethylamino)pyridine (DMAP) in 10 mL of acetonitrile. All reagent solutions including coupling agent and catalyst solution were stable for at least one week in daylight at room temperature. Synthesis of Derivatization Agent. Synthesis of 9-(2Hydroxyethyl)acridone (HEA). Acridone (25.0 g), ethylene carbonate (20.0 g), and trace amounts of KOH were dissolved together in 80 mL of N,N′-dimethylformamide in a 500-mL flask. The contents of the flask were rapidly heated to boiling. After a 6-h period, the solution was concentrated using a rotary evaporator. The residue was extracted four times with warm ether; the combined ether layers were concentrated in vacuum to yield a yellow oil liquid. The liquid was poured into a beaker containing 400 mL of water that was vigorously stirred. The mixture was filtered and the filtrate was dried at room temperature. The crude products was recrystallized twice from benzene to give the pure product, yield 12.0 g (39.5%): mp 212.2-213.8 °C; IR (KBr) 3325.8 (-OH), 1609.9 (ph-CdO), 1569 (ph), 1499.9, 1460.6, 1380.1, 1290.1, 755.9, 673.5 nm-1; m/z 239 (M+), 208 (M - CH2OH), 180 (M - NCH2CH2OH), 152, 77; 1H NMR δ 1.66-1.73 (-CH3), 5.315.52 (ph-CO), 7.08-8.08 (ph), 12.3 (OH). Found: C, 75.31; H, 5.44; N, 5.86. Calcd: C, 75.30; H, 5.45; N, 5.85. Chromatographic Method. The HPLC separation of derivatives was performed on a Hypersil BDS C18 column with a ternary gradient. Eluent A was 30% acetonitrile with 2.5% DMF, B was 70% acetonitrile with 2.5% DMF, and C was 100% acetonitrile with 2.5% DMF. The flow rate was constant at 1.0 mL/min, and the column temperature was kept at 35 °C. The fluorescence emission wavelength was set at 440 nm (excitation at 404 nm). Unless stated otherwise, the gradient condition used for the separation of amine derivatives was shown in Table 1. Sample Preparation. Five wastewater samples collected from different sampling sites were analyzed using the developed method. Once five water samples had been collected, they were immediately cooled to 0-5 °C to avoid volatilization of amines, filtered through a 0.45-µm membrance filter, and stored in glass bottles in a refrigerator. All analyses were immediately subjected to derivatization and analysis by HPLC within one day after sampling. Derivatization Procedure. Method A. To 100 µL of reagent (HEA, 4.0 × 10-5 mol/L) acetonitrile solution was successively
added 100 µL each of CDI and DMAP. This mixture was allowed to react at room temperature for 15 min to form an activated amide intermediate without being further purified. After activation of HEA, 100 µL of standard mixture of amines was added, and the tube was sealed and heated at 60 °C for 10 min. After the reaction was completed, the mixture was to be cooled at room temperature. A 10-µL volume of the crude reaction mixture was diluted to 100 µL with acetonitrile. The diluted solution (10 µL) was injected directly onto the chromatograph. The derivatization procedure is shown in Figure 1 (method A). Method B. To 100 µL of a standard mixture of amines was successively added 100 µL each of derivatizing reagent (HEA, 4.0 × 10-5 mol/L), DMAP, and CDI solution into a screw-capped tube. The mixture was allowed to react in a water bath at 60 °C for 10 min. After the reaction was completed, the mixture was cooled at room temperature. A 10-µL volume of the crude reaction mixture was diluted to 100 µL with acetonitrile. The diluted solution (10 µL) was injected directly onto the chromatograph. RESULTS AND DISCUSSION The main challenge of the present work was to develop a new fluorescence reagent and to test its feasibility in a variety of conditions. For this purpose, the fluorescent properties of the representative derivative AOCD-butylamine were tested in various solvents and temperature. As a demonstration of this study, we also measure the dependence of the derivatization procedure on reaction conditions including reaction time, temperature, coupling agent, and the catalyst utilized for the derivatization. Excitation and Emission Spectra of HEA. Acridone, which is a chromophore from the HEA molecule, has proved useful as a sensitive chromophoric moiety previously validated in our laboratory.35-37 The uncorrected excitation and emission spectra of the AOCD-butylamine derivative in methanol and acetonitrile are investigated, respectively. The results indicate that the maximal excitation and emission wavelengths are 404 and 440 nm, respectively. The profile was not shown. Stability of the AOCD-Butylamine Derivative. The stability of AOCD-butylamine derivatives at room temperature was investigated by analyzing corresponding derivatives. The relative stability of amine derivatives to I.S. (decylamine derivative) was investigated by analyzing representative derivatives (C4, C5, and C7 amines) over a period of 24 h. As expected, daylight had no effect on stability. In addition, several vials of derivatives were diluted twice with mobile phase C (70% acetonitrile containing 2.5% DMF) and were stored at 5 °C in a refrigerator. The stability of the derivatives was investigated at one-day intervals for one week; no significant change in peak area ratios of the derivatives to I.S. was found. It indicated that the derivatives were stable enough during the period of HPLC analysis. Effects of Solvents on the Fluorescence Intensity of the AOCD-Butylamine Derivative. Solvents are classified in four groups: group 1, methanol, ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol, and 1-heptanol; group 2, methylamine, ethylamine, 1-propylamine, and 1-butylamine; group 3, formic acid, glacial acetic acid, propionic acid, butyric acid, and pentanoic acid; group 4, acetonitrile/water mixture (0-100% acetonitrile, v/v). The fluorescence intensity of AOCD-butylamine derivative in solvent group 1 increases with increasing carbon numbers of alcohols (Table 2). It is probably due to the fact that hydrogen bond acting Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
263
Figure 1. Scheme for the derivatization reaction of HEA with amine compounds in the presence of N,N′-carbonyldiimidazole and 4-(dimethylamino)pyridine (base catalyst).
forces between the derivative and various alcoholic molecules decrease with increasing solvent viscosity. The fluorescence intensity in methanol is close to 1.5 times less relative to that in 1-octanol. The emission intensity of AOCD-butylamine derivative in solvent group 2 increases with increasing carbon numbers of the amines. Note that the fluorescence intensity in methylamine is also close to twice less relative to that in 1-butylamine. This is probably because the hydrogen bond acting force is stronger in methylamine than that in 1-butylamine. With the solvents from group 3, the fluorescence intensity of AOCD-butylamine increases with increasing carbon numbers of the acids, which is also similar to those as described in the cases of alcohols. The result indicates that the fluorescence intensity of AOCD-butylamine in formic acid is close to 5 times less relative to that in butyric acid. It is probably due to the fact that the derivative is partially protonated by its nitrogen atoms of molecules in relatively strong formic acid solvent, resulting in weak fluorescence emission. The emission intensities of AOCD-butylamine in various concentrations of acetonitrile were studied. The emission intensity increases with increasing acetonitrile concentration. The difference of 13.8% in fluorescence intensity is observed between 100% acetonitrile and pure water. The solvent polarity shows little effect on the emission spectra. The maximum emission remains unchanged in all cases. Effect of Temperature on Fluorescence Intensity in Various Nonaqueous Solvents. The fluorescence intensity of the fluorophore in nonaqueous solvent was investigated using varying temperatures. It is suitable for thermometry over the temperature range of 15-60 °C. Here, an excitation wavelength (λex 404 nm) 264 Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
was chosen for recording fluorescence spectra. The temperature was controlled by a thermostated curette holder connected to a circulator. It is known that the viscosity of the medium surrounding the fluorescence probe has a direct correlation between the fluorescence measurements and the bulk viscosity. This type of correlation was used to estimate the microvisicosity afforded in different organic solvents at different temperatures. In this study, It was found that AOCD-butylamine derivative is thermally stable and exhibits no significant decomposition over the temperature ranges investigated. A kinetic analysis of fluorescence intensity of AOCD-butylamine derivative at different temperature in various solvents led to a linear correlation between lnIem and 1/T. The slopes of the working curves for emission stabilization energies are shown in Table 3. Additionally, the steady-state fluorescence intensity of AOCDbutylamine derivative in aqueous solution was also investigated using varying temperatures from 15 to 60 °C. It was found that the emission intensity of AOCD-butylamine decreases with increasing temperature (diagram not shown), possibly due to the loss of excited-state energy through hydrogen bonding or due to the protonation in strong hydrogen-bonding solvents leading to a corresponding decrease in emission intensity. The effects of temperature on the fluorescence spectra of AOCD-butylamine in solvents from group 1 to group 5 (Table 3) were investigated over the temperature range from 15 to 60 °C in 15 increments. It was found that increasing the temperature does not cause any visible shift in the fluorescence spectra. Moreover, recording spectra from 15 to 60 °C allowed five groups of solvents
Table 2. Effect of the Solvent Nature on AOCD-Butylamine Spectral Propertiesa solvent group 1 methanol ethanol 1-propanol 1-butanol 1-pentanol 1-hexanol 1-heptanol 1-octanol group 2 methylamine ethylamine 1-propylamine 1-butylamine group 3 formic acid glacial acetic acid propionic acid butyric acid pentoic acid group 4 dichloromethane chloroform tetrachloromethane group 5 methyl sulfoxide tetrahydrofuran N,N-methylformamide acetonitrile
excitation (nm)
fluorescence emission (nm)
relative intensity
404 404 404 404 404 404 404 404
440 440 440 440 440 440 440 440
106.2 134.8 135.6 141.8 150.8 152.7 165.1 167.4
404 404 404 404
440 440 440 440
45.8 73.2 87.3 92.7
404 404 404 404 404
440 440 440 440 440
24.9 132.1 133.2 134.3 136.6
404 404 404
440 440
96.9 23.2 b
404 404 404 404
440 440 440 440
100c 68.2 45.7 97.4
a Concentration of AOCD-butylamine in each solvent is kept at 1.0 × 10-6 mol/L. b Emission intensity is not observed. c Relative fluorescence intensity of methyl sulfoxide as 100%.
to be distinguished in terms of changes of AOCD-butylamine emission intensity (Table 3). For solvents from group 1, AOCDbutylamine fluorescence intensity (I ) 100, at 15 °C) is remarkably decreased by elevating the temperature up to 60 °C (I ) 67-81); such a decrease is almost completely reversible when the temperature is returned from 60 to 15 °C. For solvents from group 2, AOCD-butylamine fluorescence intensity is also largely decreased (I ) 68-76 at 60 °C) but still excellently reversible; for the solvents from group 3, a large decrease in emission intensity is also observed (I ) 67-69 at 60 °C), but the formic acid effect is irreversible; for solvents from group 4 (aprotic solvent), the decrease is slight (I ) 87-90 at 45 °C; the solvents from group 4 are easily volatilized, so the investigated temperature range was from 15 to 45 °C) and irreversible for dichloromethane. In addition, it was also found that with tetrachloromethane, in which the emission intensity of AOCD-butylamine is completely quenched, however, increasing the temperature does not cause any visible shift in the fluorescence spectra; for solvents from group 5, the decreases of AOCD-butylamine fluorescence intensity exhibit no remarkable differences as observed above. The result is that the decrease in emission intensity at 60 °C in solvents from methyl sulfoxide, tetrahydrofuran, and N,N-dimethylformamide shows similar results (I ) 75 at 60 °C). Derivatization Conditions. Amino compounds react with AOCD intermediate prepared by the reaction of HEA with coupling agent (CDI) to give the sensitively fluorescent derivatives. To ensure maximum derivatization of the amines, reaction
conditions were investigated for optimizing derivatization with butylamine as noted below. Effects of Solvent and Base Catalyst on Derivatization. Acetonitrile, ethyl acetate, acetone, chloroform, and dichloromethane were investigated as reaction cosolvents for the derivatization of amines. The results indicated that acetonitrile and ethyl acetate gave the best results as assessed by the detector response (Table 4). A remarkable decrease in detector responses in chloroform (46%), ethyl acetate (93%), and dichloromethane solvents (75%) was observed, respectively. Acetone gave the lowest response (40%) under the conditions proposed. In general, acetonitrile was used as the reaction solvent throughout this study because of the immiscibility of ethyl acetate in the mobile phase. Several base catalysts, including triethylamine, pyridine, 2-methylpyridine, 2,4-dimethylpyridine, 2,6-dimethylpyridine, 2,5-dimethylpyridine, and DMAP, were evaluated as reaction catalysts for the formation of the activated amide intermediate in the derivatization procedure. Reactions were carried out at 60 °C for 10 min with 50 ng/mL amines (C4, C5, and C7) using CDI as the coupling agent in the presence of various base catalysts (0-40 mg/mL). Each value is an average of six runs with the detector response obtained with DMAP taken as 100%. The results (Table 5) indicate that DMAP is the best base catalyst and gives the highest detector responses for the derivatization. Further study about the effect of the added amount of DMAP (>20 mg/mL) on the derivatization was tested. The results indicate that the added amount of DMAP in excess of 20 mg/mL does not significantly increase the reaction yield. In view of this, all derivatization is carried out by the use of DMAP (20 mg/mL) as the base catalyst. Effects of the Concentration of Coupling Agent on Detector Response. Derivatization with ′CDI as reaction coupling agent was investigated. The effect of the CDI concentration on the peak height is shown in Figure 2. Relatively higher detection responses were obtained in the concentration range 5-10 mg/ mL (corresponding 3.0 × 10-3-6.0 × 10-3 mol/L) CDI without affecting the impurity peaks, and thus, 10 mg/mL coupling agent concentration was selected in subsequent studies. Temperature Conditions and Time Effects for Derivatization. Temperature is a very important factor in optimizing the derivatization rate. Investigation of the effect of the temperature on the formation of the amine derivatives showed that the derivatization rate gradually increased with increasing temperature. When tested at different temperatures over various periods of time, the reaction was completed within 6, 8, and 10 min at 100, 80, and 60 °C, respectively. The peak heights for all amines become constant at 10 min at 60 °C, but the produced derivatives slightly decreased with reproducible quantitative yields when the temperature was over 80 °C (Figure 3). It was possible that above 80 °C the position of the equilibrium reduced the proportion of AOCD reacting with amines because of forming some byproducts (unidentified); while below 60 °C the rate of reaction was decreased and led to a long derivatization time. On the other hand, a longer reaction time (at 60 °C, >30 min) led to a slight decrease in signal for all derivatives. Therefore, for most subsequent derivatization, the temperature selected in this study was 60 °C. Increasing the reagent concentration to more than 4.0 × 10-5 mol/L did not significantly alter the reaction time and temperature needed for the derivatization reaction to be completed. Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
265
Table 3. Effects of Temperature on AOCD-Butylamine Fluorescence Intensity in Various Solvents intensity solvent group 1 methanol ethanol 1-propanol 1-butanol 1-pentanol 1-hexanol 1-heptanol 1-octanol group 2 methylamine ethylamine 1-propylamine 1-butylamine group 3 formic acid glacial acetic acid propionic acid butyric acid pentoic acid group 4 dichloromethane chloroform tetrachloromethane group 5 methyl sulfoxide tetrahydrofuran N,N-methylformamide acetonitrile a
15 °C
30 °C
45 °C
60 °C
stability energy (kJ/mol)
regression coefficient
100 100 100 100 100 100 100 100
91 91 92 91 90 92 93 89
83 82 84 82 81 83 87 77
75 73 77 73 72 74 81 67
5.08 5.57 4.71 5.57 5.81 5.34 3.72 7.14
0.9978 0.9956 0.9983 0.9956 0.9975 0.9937 0.9989 0.9963
100 100 100 100
88 89 90
78 79 83
68 71 76 a
6.79 6.09 4.82
0.9979 0.9996 0.9994
100 100 100 100 100
87 87 86 85 84
78 79 78 76 75
69 69 68 67 67
52 6.44 6.69 6.99 7.04
0.9994 0.9972 0.9972 0.9986 0.9973
100 100
95 94
90 87
a a b
2.69 3.55
0.9991 0.9954
100 100 100 100
95 94 90 91
97 87 81 82
75 75 75 79
5.02 4.97 5.17 5.03
0.9991 0.9954 0.9993 0.9985
No determination. b Emission intensity is not observed.
Table 4. Effect of Solvent on Detector Response for the Derivatization Reactiona
Table 5. Effect of Catalyst on the Yield (Detector Response) of the Derivatizationa
solvent
detector response (%)
base catalyst (40 mg/mL)
detector response (%)
ethyl acetate acetone acetonitrile chloroform dichloromethane
93 40 100 46 75
triethylamine pyridine 2-methylpyridine 2,4-dimethylpyridine 2,6-dimethylpyridine 2,5-dimethylpyridine 4-(dimethylamino)pyridine no catalyst
33 64 50 23 36 30 100 20
a Reactions were carried out at 60 °C for 30 min with 50 ng/mL C , 4 C5, and C7 amines. CDI and DMAP concentrations were 10 and 20 mg/mL, respectively. Each value is an average of six runs with the detector responses obtained with acetonitrile taken as 100%.
Analytical Performance. The calibration graph was carried out by injecting 10-µL volumes of solutions containing known amounts of AOCD-heptylamine derivative equivalent to 10, 20, 40, and 50 ng/mL heptylamine containing the internal standard AOCD-decylamine. The calibration graph was established with the peak height ratio (y) of derivatized heptylamine to AOCDdecylamine (internal standard) versus heptylamine concentration (x); the linear regression equation obtained was y ) 0.0213 + 0.0264x (n ) 6, γ2 ) 0.998). The precision and reproducibility of the proposed method were investigated for six replicates by the analysis of heptylamine at 15, 25, and 40 ng/mL with the integration peak area. The results are shown in Table 6. RSDs of intraday and interday values at the three concentrations are all below 4.5%. These coefficients of variation can be considered acceptable if one takes into account the matrix complexity. Linear dynamic ranges and detection limits for amines using this method 266 Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
a Reactions were carried out at 60 °C for 10 min with 10 ng/mL C , 4 C5, and C7 amines using CDI as the coupling agent (10 mg/mL) in the presence of the various types of base catalysts. Each value is an average of six runs with the detector responses obtained with 4-(dimethylamino)pyridine taken as 100%.
are listed in Table 7. As expected, amines exhibited wider linearity ranges. In this study, even lower detection limits should be possible, but the practically obtained values are limited due to the purity of the reagents. For the separation of derivatized amines, a BDS C18 column eluted with various mobile-phase compositions (acetonitrile or methanol in water) in combination with gradient elution was tested. For the simultaneous separation of amine derivatives, two types of mobile-phase compositions gave the best separation profiles with the shortest retention time. (1) Eluent: A 30% acetonitrile with 2.5% DMF; B 70% acetonitrile with 2.5% DMF; C 100% acetonitrile with 2.5% DMF (elution conditions as method 1 in Table 1). This elution system gives the sharpest peaks and short retention time (Figure 4). (2) Eluent: A 40% acetonitrile
Table 7. Linear Dynamic Ranges and Procedural LODs of Several Representative Amines Determined Using Derivatization Method A compound
linear range (ng/L)
RSDa (%)
LODsb (ng /mL)
methylamine ethylamine isoamylamine heptylamine undecylamine
23-810 22-715 26-675 36-816 65-785
4.0 3.6 3.0 4.0 4.5
0.16 0.18 3.4 7.5 12.8
a Relative standard deviation determined for 50 ng/mL of each AOCD derivatives. bSignal-to-noise ratio, 3.
Figure 2. Effect of concentration of CDI on derivatization. A 50 ng/ mL butylamine solution was tested by the derivatization procedure using various concentrations of CDI.
Figure 3. Dependence of detector response and time of completion of the derivatization reaction on temperature. Reactions were carried out at the specified temperature with 50 ng/mL heptylamine in the presence of HEA, CDI, and DMAP, at concentrations were 4.0 × 10-5 mol/L, 10 mg/mL, and 20 mg/mL, respectively. Table 6. Precision and Accuracy for the Analysis of Heptylamine (n ) 6) concentration (ng/mL)
intraday interday
theoretical
found
RSD (%)
recovery (%)
15 25 40 15 25 40
15.7 24.4 38.4 14.7 24.6 38.8
4.4 3.0 3.1 4.2 3.7 4.0
104.6 97.6 96.0 98.0 98.4 97.0
and B 100% acetonitrile solution (eluted conditions as method 2 in Table 1) give the short retention time, but all peaks lead to broad peak shapes (Figure 5). It is probably due to the fact that the addition of DMF into the mobile phase to avoid the absorption between derivative molecules and residually active sites onto the surface of the padding. Unless stated otherwise, mobile-phase compositions and gradient conditions from method 1 in Table 1
Figure 4. HPLC chromatogram of a standard mixture of 50 ng/mL standard derivatives of amines detected by fluorescence monitor (excitation wavelength 404 nm, emission 440 nm). Chromatographic conditions: column 200 × 4.6 mm Hypersil BDS C18 (5 µm); eluent A, 30% acetonitrile with 2.5% DMF; eluent B, 70% acetonitrile with 2.5% DMF; eluent C, 100% acetonitrile with 2.5% DMF; flow rate 1.0 mL/min with gradient as Table 1 (method 1); column temperature 35 °C; Peaks: C1, methylamine; C2, ethylamine; iso-C3, isopropylamine; C4, butylamine; C5, amylamine; C7, heptylamine; C10, decylamine; C11, undecylamine; C14, tetradecylamine.
was selected in all cases. However, discrimination between Figure 4 obtained with derivatization method A and Figure 5 obtained with derivatization method B for the detection of amines based on the fluorescence responses was significantly observed. It was probably because the derivatization in method A was a rapid nucleophilic substitution reaction in the presence of a base catalyst. In method B (here, reagent (HEA), coupling agent (CDI) and amine molecules added simultaneously in solution), it is possible that alkylamines are superior to a hydroxyl group from HEA molecules when reacting with ′CDI This procedure probably forms another intermediate N-alkylcarbonylimidazole (ACDI, see Figure 1). Subsequently, the nucleophilic substitution reaction of Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
267
Figure 5. Chromatogram of a standard mixture of 50 ng/mL standard derivatives of amines detected by fluorescence monitor (excitation wavelength 404 nm, emission 440 nm). Chromatographic conditions: column 200 × 4.6 mm Hypersil BDS C18 (5 µm); eluent A, 40% acetonitrile; eluent B, 100% acetonitrile; flow rate 1.0 mL/ min. Gradient condition as in Table 1 (method 2). Column temperature 35 °C; Peaks as in Figure 4.
Figure 6. Relative reactivities (reactivities are normalized to ethylamine as 100%) of amines in the derivatization under the conditions proposed in experiment. EYA, ethylamine; DMYA, dimethylamine (EYA/DMYA ) 40.0); PYA, propylamine; IPYA, isopropylamine (PYA/ IPYA ) 4.2); BYA, butylamine; IBYA, isobutylamine (BYA/IBYA ) 4.32); DEYA, diethylamine (BYA/DEYA ) 118.7).
ACDI with HEA molecules is a slow reaction under the reaction conditions proposed owing to the alkalescence of hydroxyl groups from HEA molecules, which results in low detection responses for the derivatization. Therefore, the derivatization method A was selected in all cases in this study. A greater degree of deviation for the detection response is observed in the comparison of propylamine and isopropylamine, where the latter is ∼4.22 times less reactive than the former. The same result was also observed for butylamine (BYA) and isobutylamine (IBYA) (the ratio of BYA/IBYA is 4.32 (Figure 6)). The common trait of this type of nucleophilic substitution reaction is steric requirements for solute molecules. Diethylamine or isopropylamine imposes two alkyl groups around the nitrogen atom, which can assume many more spatially demanding configurations than those of propylamine or 268 Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
Figure 7. A typical HPLC profile of wastewater from a paper mill (Shandong, Qufu, China). Chromatographic conditions and peaks as Figure 4. Peaks: C1, methylamine; C2, ethylamine; E, diethylamine; iso-C3, isopropylamine; C3, propylamine; iso-C4, isobutylamine; C4, butylamine; iso-C5, isoamylamine; C5, amylamine; C6, hexylamine; C7, heptylamine (internal standard); C, 1,5-pentanediamine; A and B (unidentified).
butylamine for a nucleophilic substitution reaction. The detector response for dimethylamine was ∼50 times less reactive than ethylamine. The reactivity for 1-butylamine was more than 1 order of magnitude higher than diethylamine under the same experimental conditions. Application to Real Samples. Figure 7 shows the application of this method for the determination of amine contaminants present in wastewater samples collected from several different sampling sites. The contents of amines from several different sampling sites are listed in Table 8. The highest concentrations of amine contaminants were detected from a paper mill (Shandong, Qufu, China) serving larger populations. It is obvious that the highest concentration of contaminant was isoamylamine. Generally, the amine concentrations depended on the type of wastewater samples. While methylamine and ethylamine concentrations remain the same in domestic wastewater and melted snow water, which is consistent with low concentration. The contents of >C4 amines in domestic wastewater and melted snow water were, respectively, almost undetected except that 1,5-pentanediamine was determined in domestic wastewater. CONCLUSION Although amine compound analysis is a relatively mature technique, current liquid chromatographic technology still needs improvement in the speed, accuracy, and sensitivity of the analysis. In our study, HEA was first activated to its amide intermediate in the presence of coupling agent (CDI); which was not only sensitive for fluorescent detection but also a simple and convenient derivatization method for the determination of amines including short-, medium-, and long-chain amines. The proposed method
Table 8. Analysis of Amines from Five Wastewater Samples (µg/L)a sample no. compound
1
2
3
4
methylamine ethylamine diethylamine isopropylamine propylamine isobutylamine butylamine isoamylamine amylamine hexylamine heptylaminec 1,5-pentanediamine
161 124 46 108 184 4.6 24 246 18 nd 76 38.7
74 63 8.0
43 21 4.8
C15), in fact, can easily be achieved under the HPLC derivatization conditions proposed. The proposed method has been successfully applied to the determination of free amines in real wastewater samples with satisfactory results. A possible disadvantage of this method is that the reagent HEA or amide intermediate AOCD can only be used in the precolumn mode. Further work will also seek to improve the experimental conditions and explore the trace-level application. ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation under Grant 20075016.
Received for review March 12, 2001. Accepted August 29, 2001. AC010285D
Analytical Chemistry, Vol. 74, No. 1, January 1, 2002
269
