Identification of gas emanated from human skin: methane, ethylene, and ethane.

We investigated whether methane, ethylene and ethane gas can be detected in gas emanating from human skin, which is called skin gas. Skin gas was collected with a homemade stainless-steel trap system, which was cooled with liquid nitrogen, and analyzed with a gas chromatograph fitted with a flame ionization detector (FID). Skin-gas samples were obtained by covering a hand for 30 min with a polyfluorovinyl bag in which pure helium gas was introduced. The bag, the trap system and GC were set up online to avoid any contamination by air. Methane, ethylene and ethane in skin gas were successfully collected at an average amount emanated for 30 min (from ten subjects) of 150 +/- 63, 20 +/- 11 and 17 +/- 8 [mean +/- SD] pg/cm2, respectively.     

Continuous-flow enzymatic determination of creatinine with improved on-line removal of endogenous ammonia

A new continuous-flow automated enzymatic method suitable for the direct determination of creatinine in physiological samples is described. The proposed system utilizes an on-line gas predialysis unit in conjuction with a flow-through enzyme reactor coil and a potentiometric ammonia detector. The enzyme reactor contains immobilized creatinine iminohydrolase (EC 3.5.4.21) which converts creatinine to ammonia and N-methylhydantoin. Ammonia liberated from this reaction is detected downstream with the membrane electrode-based detector. The novel gas predialysis unit effectively removes >99.8% of endogenous ammonia (up to 1 mM) present in the sample. Thus, final peak potentials recorded by the electrode detector are directly proportional to the logarithm of creatinine concentrations present. The method is shown to be precise (<3%), selective, and capable of accurately determining creatinine in serum and urine samples containing abnormally high endogenous ammonia levels. Determinations of creatinine in serum samples (n = 30) using this new method correlate well with an existing Technicon AutoAnalyzer colorimetric method (r = 0.996).     

Chemical composition and in vitro antimicrobial activity of the essential oil of the flowers of Tridax procumbens

The essential oil of the flowers of Tridax procumbens L. was obtained by hydro-distillation and analyzed by gas chromatography equipped with a flame ionization detector (GC-FID) and gas chromatography coupled with mass spectrometry (GC/MS). Twenty-six compounds were identified, which comprised 90.6% of the total constituents. The most abundant compound was (Z)-falcarinol (25.9%), followed by alpha-selinene (15.3%), limonene (8.3%) and zerumbone (4.3%). Antimicrobial activity was tested against six Gram-positive and eight Gram-negative bacteria, and three fungi. The oil was active against the tested Gram-positive bacteria at a concentration range of 0.14 +/- 0.03 - 0.57 +/- 0.05 mg/mL, while 0.67 +/- 0.12 - 4.58 +/- 0.41 mg/mL was effective against the studied Gram-negative bacteria. Remarkable antifungal activity was found against the tested fungi at a concentration range of 0.06 +/- 0.008 - 0.10 +/- 0.01 mg/mL.     

Miniaturized Optical System for Detection of Ammonia Nitrogen in Water Based on Gas-Phase Colorimetry

A small-size gas-tight optical measuring system for detection of ammonia nitrogen in water was prepared based on gas-phase ammonia induced color change of the sensing element that was made by loading bromothymol blue (BTB) in a transparent porous glass fiber membrane. The gas-tight optical measuring system consists of a gas-testing and a liquid-sample chamber connected with each other by means of tubes and a mini-pump that cycles the gas between the two chambers. A 625-nm light emitting diode (LED), a photodetector and a sensing element were mounted in the gas-testing chamber for optical response to ammonia gas released from the water in the liquid-sample chamber. Release of ammonia gas was realized by alkalinizing the water sample with NaOH. Owing to the amount accumulation of ammonia gas in the sealing system, the ammonia nitrogen detection limit of the device can be very low. A small concentration of ammonia nitrogen, as low as 0.05 mg/L, was detected. The two linear-response ranges from 0.05 mg/L to 0.26 mg/L and from 0.26 mg/L to 2.62 mg/L were obtained. A relative standard deviation of ≤1% was determined by multiple measurements of the same sample.     

Effect of ammonia carrier gas on the response of the flame ionization detector

The effect of using ammonia as a carrier gas on the response of the flame ionization detector (FID) has been investigated. It was found that the FID response, calculated as the effective carbon number (ECN), increased for all the compounds studied when ammonia, rather than helium, was used. The change was 0–0. 9 carbon atom for hydrocarbons, one carbon atom for alcohols and diphenyl ether, and 0.4–1 carbon atom for phenols and ketones. The increase in ECN was larger for amines (0. 8–5 carbon atoms), but these numbers also reflected an improvement in chromatographic performance as a result of reduced adsorption on the column. The largest change in signal-to-noise ratio, a six-fold increase, was obtained for octyl-amine; ratios for hexyl methyl ketone, diisobutyl ketone, dihexyl-amine, dibutylamine, and N-methyloctylamine increased by a factor of 2–3 when ammonia was used as carrier gas. To determine the extent to which the effect on detector response was solely attributable to ammonia, a mixture of 5 % ammonia in nitrogen was used as detector make-up gas with helium as carrier gas. Under these conditions the noise in the FID increased but for most of the compounds studied the signal-to-noise ratio also increased.     

Study by means of high-performance liquid chromatography of solutes that decrease theophylline/protein binding in the serum of uremic patients.

Substantial changes in protein binding of drugs occur during the progression of renal insufficiency. Protein-bound uremic solutes play a role in the inhibition of drug protein binding. We previously demonstrated that hippuric acid in uremic ultrafiltrate was an inhibitor of the theophylline protein binding. The present study was undertaken to extend the yield of protein-bound uremic solutes by displacing ligands in uremic serum from their binding sites by five deproteinization methods. The inhibitory effect on theophylline protein binding of the deproteinized uremic serum was higher than with ultrafiltrate (p < 0.05). The influence of 30 semi-preparative HPLC fractions from deproteinized uremic serum on the theophylline protein binding was evaluated to identify the responsible compounds and to compare their relative individual impact. The theophylline protein binding was calculated as a percentage (bound versus total). The most important decrease of the protein binding was observed in HPLC fractions 6, 10 to 13, 15 and 28 with protein binding of: 61.5 +/- 10.8, 64.5 +/- 7.6, 60.9 +/- 10.1, 47.5 +/- 3.3, 60.0 +/- 6.7, 60.7 +/- 6.3 and 61.3 +/- 6.9%, respectively versus 69.1 +/- 2.4% for control serum (p < 0.05). The responsible compounds were characterized in the fractions by co-elution: 3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid (CMPF), indole-3-acetic acid, indoxyl sulfate, hippuric acid, p-hydroxyhippuric acid and tryptophan. Their concentration was determined by analytical HPLC and a solution containing these compounds at the same concentration as in deproteinized uremic serum was composed. This solution was added to control serum and decreased the theophylline protein binding from 69.0 +/- 4.4% to 61.3 +/- 1.3%, which was less important than in genuine uremic serum (44.4 +/- 3.8%, p < 0.05). Dose-response curves with the characterized compounds revealed that the most important role in binding inhibition could be attributed to hippuric acid and CMPF. Our data suggests that the yield of protein binding inhibiting compounds is more important with deproteinized uremic serum than with uremic ultrafiltrate. The identified uremic compounds are not entirely representative for the decreased protein binding of theophylline, indicating that additional factors than those identified in this study affect the protein binding as well.     

Objective measurement of minimal erythema and melanogenic doses using natural and solar-simulated light

Very little information exists on the amount of natural and artificial UV light required to cause sunburn and tanning in individuals with very pale skin who are at the greatest risk of developing skin cancer. We have investigated minimal erythema dose (MED) and minimal melanogenic dose (MMD) in a group of 31 volunteers with Fitzpatrick skin types I and II using an Oriel 1000 W xenon arc solar simulator and natural sunlight in Sydney, Australia. We measured the erythemal and melanogenic responses using conventional visual scoring, a chromameter and an erythema meter. We found that the average MED measured visually using the artificial UV source was 68.7 +/- 3.3 mJ/cm2 (3.4 +/- 0.2 standard erythema doses [SED]), which was significantly different from the MED of sunlight, which was 93.6 +/- 5.6 mJ/cm2 (P < 0.001) (11.7 +/- 0.7 SED). We also found significant correlations between the solar-simulated MED values, the melanin index (erythema meter) and the L* function (chromameter). The average MMD (obtained in 16 volunteers only) using solar-simulated light was 85.6 +/- 4.9 mJ/cm2, which was significantly less than that measured with natural sunlight (118.3 +/- 8.6 mJ/cm2; P < 0.05). We mathematically modeled the data for both the chromameter and the erythema meter to see if we were able to obtain a more objective measure of MED and differentiation between skin types. Using this model, we were able to detect erythemal responses using the erythema index function of the erythema meter and the a* function of the chromameter at lower UV doses than either the standard visual or COLIPA methods.     

Effect of ammonia on the response of the electron capture detector (ECD)

The effect of ammonia on the electron capture detector (ECD) response has been investigated. Nitrogen with different ammonia concentrations (5–20%) was used as make-up gas. Compared to pure nitrogen, the ECD response decreased when the ammonia concentration in the make-up gas was 5%. However, the response increased when the ammonia concentration was 20%. The response factor of 4-chlorophenol increased 4 fold when ammonia was 20%. Also, di- and tri-chlorophenols increased by 30–50%. The nitroaromatic compounds responses increased by about 2–3.7 times with 20% ammonia in the make-up gas. The signal-to-noise (S/N) increased when 20% ammonia in nitrogen was used as make-up gas compared to pure nitrogen. Also, the detector linearity increased by 50% with ammonia.     

Iodine status and the effect of soil erosion on trace elements in Nanka and Oba towns of Anambra State, Nigeria

Iodine Deficiency Disorders (IDD) is common in all populations. Iodine and other trace elements naturally occur in the soil but erosion leaches off these elements from the soil. This results in a continued loss of trace elements from the soil. In the present study, the levels of iodine, selenium, zinc and lead in the environment (measured in soil, bitter leaves (Vernonia amygdalina), cassava roots (mannihot utilissima, staple food in Nigeria), and drinking water) and urinary iodine from school children (n=200), pregnant women (n=60) and women of child bearing age (n=60) were determined for Nanka prone to soil erosion and Oba all in Anambra State, Nigeria (used as control) to assess their risk to IDD. The levels of selenium, zinc and lead were analysed using Atomic Absorption Spectrophotometry while the levels of iodine in the environment and urinary iodine were estimated using the method of Dunn et al.,(1993). In this study there was a positive correlation between iodine and the metals. The results show that the mean concentrations of total soil zinc (0.69 +/- 0.16 ppm); lead (0.40 +/- 0.12 ppm) values in Oba were significantly (p < 0.05) higher than values from Nanka (Zn = 0.33 +/- 0.10 ppm; Pb = 0.21 +/- 0.09 ppm). However, total soil values for selenium and iodine in soil were not significantly different in the two communities. Mean concentration of total vegetable zinc (0.63 +/- 0.14 ppm) value in Oba is significantly (p < 0.05) higher than the value from Nanka (Zn = 0.31 +/- 0.07 ppm). However, total vegetable values for I, Se and Pb were not significantly different in the two communities. Also, mean concentration of total cassava zinc (0.65 = 0.15 ppm) in Oba was significantly (p < 0.05) higher than Zn (0.44 +/- 0.1l ppm) from Nanka. However, values for Se, Pb, and I were not significantly different in the two communities. Mean concentration of total water iodine (105.25 +/- 10.44 microg/L) in Oba was significantly (p < 0.05) higher than the value from Nanka (I = 89.8 +/- 6.42 microg/L). However, total water values for Se, Zn, and Pb were not significantly different in the two communities. The mean urinary iodine concentration of 170.65 +/- 27.17 microg/L in school children from Oba was significantly higher (p < 0.05) than the mean concentration of 156.12 +/- 16.48 microg/L found in school children from Nanka. However, the mean urinary iodine concentration of all the women (pregnant and non-pregnant) were not significantly different in the two communities but they are below the recommended daily intake. The results show that people living in Nanka and Oba, could be at risk of IDD.     

Electronic properties of liquid ammonia: a sequential molecular dynamics/quantum mechanics approach

The electronic properties of liquid ammonia are investigated by a sequential molecular dynamics/quantum mechanics approach. Quantum mechanics calculations for the liquid phase are based on a reparametrized hybrid exchange-correlation functional that reproduces the electronic properties of ammonia clusters [(NH3)n; n=1-5]. For these small clusters, electron binding energies based on Green's function or electron propagator theory, coupled cluster with single, double, and perturbative triple excitations, and density functional theory (DFT) are compared. Reparametrized DFT results for the dipole moment, electron binding energies, and electronic density of states of liquid ammonia are reported. The calculated average dipole moment of liquid ammonia (2.05+/-0.09 D) corresponds to an increase of 27% compared to the gas phase value and it is 0.23 D above a prediction based on a polarizable model of liquid ammonia [Deng et al., J. Chem. Phys. 100, 7590 (1994)]. Our estimate for the ionization potential of liquid ammonia is 9.74+/-0.73 eV, which is approximately 1.0 eV below the gas phase value for the isolated molecule. The theoretical vertical electron affinity of liquid ammonia is predicted as 0.16+/-0.22 eV, in good agreement with the experimental result for the location of the bottom of the conduction band (-V 0=0.2 eV). Vertical ionization potentials and electron affinities correlate with the total dipole moment of ammonia aggregates.     

Potentiometric determination of proteins with an ammonia sensor

A flow system is described for the cleavage of proteins with immobilized protease enzyme to L-amino acids which are then converted to ammonia with glass-immobilized L-amino acid oxidase. An ammonia gas electrode is used as detector. Immobilization techniques are discussed, as are optimum conditions for L-phenylalanine. Bovine serum albumin was determined in the range 0.1–100 μg ml^-1. Human blood sera required dilution for analysis.     

Sampling small volumes of ambient ammonia using a miniaturized gas sampler

The development of a gas sampler for a miniaturized ambient ammonia detector is described. A micromachined channel system is realized in glass and silicon using powder blasting and anodic bonding. The analyte gas is directly mixed with purified water, dissolving the ammonia that will dissociate into ammonium ions. Carrier gas bubbles are subsequently removed from the liquid stream through a venting hole sealed with a microporous water repellent PTFE membrane. A flow restrictor is placed at the outlet of the sampler to create a small overpressure underneath the membrane, enabling the gas to leave through the membrane. Experiments with a gas flow of 1 ml min(-1), containing ammonia concentrations ranging from 9.4 ppm to 0.6 ppm in a nitrogen carrier flow have been carried out, at a water flow of 20 microl min(-1). The ammonium concentration in the sample solution is measured with an electrolyte conductivity detector. The measured values correspond with the concentration calculated from the initial ammonia concentration in the analyte gas, the fifty times concentration enhancement due to the gas-liquid volume difference and the theoretical dissociation equilibrium as a function of the resulting pH.     

Nitrogen isotope analysis of ammonium in aqueous solutions using a perfluorosulfonated ionomer membrane for solid-phase microextraction

The use of custom-made solid-phase microextraction (SPME) fibers coated with a perfluorosulfonated ionomer, Nafion, was investigated for nitrogen isotopic analysis of ammonium in aqueous solutions. Aqueous ammonium was converted to ammonia by addition of a base, followed by absorption from the headspace, desorption in the injection port of a gas chromatograph, and analysis by gas chromatography/combustion/isotope ratio mass spectrometry (GC/C/IRMS). Fibers coated with a Nafion tubing were chosen due to a higher fiber-gas distribution constant and a higher Nafion thickness than fibers coated with Nafion solution, both leading to a higher amount of ammonia absorbed at equilibrium. The Nafion membrane-coated fiber absorbed approximately 20 times more than a commercial polydimethylsiloxane (PDMS) fiber. The isotopic fractionation between fiber and gas was 1.0117 +/- 0.0009 (standard deviation, SD, of all measurements) at an initial ammonia gas concentration of 21-210 microM. At 390 microM initial gas concentration it was slightly lower. When sampling from liquid samples, an ammonium concentration of 10 mM was needed to obtain a sufficient amount of ammonia absorbed. Modeling of the absorption at different temperatures showed that the absorption was approximately constant in the temperature range suitable for SPME experiments. Absorption at room temperature was therefore used for simplicity. A pilot study was conducted in which absorption was achieved from a single 9 microL droplet of sample. The preliminary results showed that delta(15)N analysis was possible for only 0.4-0.5 micromol of ammonium with a SD of 0.8 per thousand (n = 5).     

Molecular electron affinities and the calculation of the temperature dependence of the electron-capture detector response

The use of the electron-capture detector (ECD) to measure molecular electron affinities and kinetic parameters for reactions of thermal electrons is reviewed. The advances of the past decade are emphasized and include the multistate electron-capture detector model and the use of semi-empirical self-consistent field quantum mechanical calculations and half wave reduction potential values to support gas phase experimental results. A procedure for the evaluation of the adiabatic electron affinities of the main group elements and the homonuclear diatomic molecules is presented. Potential excited states are identified for the magnetron (MGN) values for quinones, thermal charge transfer (TCT) values for CS2, C6F6, SF6 and photoelectron spectroscopy (PES) values for O2, NO, nitromethane, and the nucleic acids. Literature electron affinities are then evaluated. The temperature dependence of the electron-capture detector can be calculated using values for kinetic rate constants and electron affinities to optimize response and temperature sensitivity in analytical procedures. The temperature dependence for adenine, guanine, thymine and cytosine are predicted for reactions with thermal electrons. Using the recent advances, the new adiabatic electron affinities are: all in electron volts (eV), 4-F-benzaldehyde (0.57 +/- 0.05) and acetophenones (APs) 4-F-AP (0.52 +/- 0.05); 2-CF3-AP (0.79 +/- 0.05); 3-CF3-AP (0.79 +/- 0.05); 4-CF3-AP (0.89 +/- 0.05); 3-CI-AP (0.67 +/- 0.05); and 4-Cl-AP (0.64 +/- 0.05). The adiabatic electron affinities of chloro and fluorobenzenes range from 0.17 to 1.15 eV and 0.13 to 0.86 eV.     

Application of copper sulfate pentahydrate as an ammonia removal reagent for the determination of trace impurities in ammonia by gas chromatography

Rapid analysis of trace permanent gas impurities in high purity ammonia gas for the microelectronics industry is described, using a gas chromatograph equipped with a phtoionization detector. Our system incorporates a reactive precolumn in combination with the analytical column to remove the ammonia matrix peak that otherwise would complicate the measurements due to baseline fluctuations and loss of analytes. The performance of 21 precolumn candidate materials was evaluated. Copper sulfate pentahydrate (CuSO₄·5H₂O) was shown to selectively react with ammonia at room temperature and atmospheric column pressures, without affecting the hydrogen, oxygen, nitrogen, methane or carbon monoxide peak areas. To prevent loss of trace carbon dioxide, an additional boron trioxide reactant layer was inserted above the copper sulfate pentahydrate bed in the reactive precolumn. Using the combined materials, calibration curves for carbon dioxide proved to be equivalent in both ammonia and helium matrix gases. These curves were equivalent in both matrix gases. The quantitative performance of the system was also evaluated. Peak repeatabilities, based on eight injections, were in the range of 4.1–8.2% relative standard deviation; and detection limits were 6.9 ppb for H₂, 1.8 ppb for O₂, 1.6 ppb for N₂, 6.4 ppb for CH₄, 13 ppb for CO, and 5.4 ppb for CO₂.     

Peculiarities of gas chromatographic analysis of 6N4 volatile hydrides using a helium ionization detector operated in the ionization amplification mode

A pneumovacuum circuit of a Tsvet-800-2 gas chromatograph with a helium ionization detector (HID) in the ionization amplification mode is proposed for controlling limiting trace impurities in ammonia, arsine, monosilane, and phosphine of grade 6N4. The conditions for the preparation and reliable operation of the chromatograph with the HID are systematized. The effect of an additive of (0.0–17.6) × 10^–4 mol % of hydrogen in the carrier gas of the detector on the polarity of its signal on the impurities of oxygen and nitrogen at a level of (0.5–20.0) × 10^–5 mol % is studied. Methods are proposed for eliminating systematic errors in the determination of the concentration of oxygen and nitrogen in the range from 0.5 × 10^–5 to 1 × 10^–3 mol % in 6N4 volatile hydrides, caused by “counter” impurities in the carrier gas, the mechanism of the heteropolar sensitivity of the HID, and irreversible chemical reactions of oxygen traces with phosphine and products of its decomposition on a CaA–ShM zeolite in a separating column of a chromatograph. Gas chromatography analysis of ammonia, arsine, monosilane, and phosphine of grade 6N4 was carried out.     

Analysis of low levels of oxygen, carbon monoxide, and carbon dioxide in polyolefin feed streams using a pulsed discharge detector and two PLOT columns

A gas chromatography (GC) method is developed for rapid analysis of polyolefin feed streams for the catalyst poisons CO, CO(2), and O(2). The method uses an HP MoleSieve column in parallel with a CP-PoraPLOT Q column and a pulsed discharge detector (PDD). Detection limits for each of the potential poisons are between 50 and 250 ppb. For a 10-ppm standard, the precision of the method was +/- 4.2% for oxygen, +/- 7.8% for carbon dioxide, and +/- 2.0% for carbon monoxide. In addition to the polyolefin feed stream, nitrogen and hydrogen feed streams are also analyzed. In each case, sampling is observed to be a critical issue, with air contamination of the sample cylinder often the limiting step in determining the true level of oxygen. It is also noted that large amounts of argon are present in the standards when nitrogen is used as a balance gas. Because the trace oxygen peak partly coelutes with the larger argon peak, it is suggested that helium be used as the balance gas for all standards. This general experimental arrangement should be effective when applied to feed streams for other polymers as well.     

Determination of ligustilide in rat blood and tissues by capillary gas chromatography/mass spectrometry

A gas chromatography/mass spectrometry (GC/MS) method was developed to study the pharmacokinetics of ligustilide following oral administration to rats. The method was used for the analysis of samples taken from rats. Biological samples were prepared by liquid-liquid extraction (LLE) using an n-hexane-ether (2:1) solvent mixture for a sample clean-up step and analyzed by GC/MS with a quadrupole MS detector in selected ion monitoring mode (m/z 190). The calibration curves were linear over the concentration range 0.172-8.60 microg/mL (r > 0.99) for blood samples and a different range (r > 0.99) for different tissue samples. The limit of detection (LOD) was 1.0 ng/mL or 1.0 ng/g (three times the signal-noise ratio). Within- and between-day precision expressed as the relative standard deviation (RSD) for the method was 1.58-3.88 and 2.99-4.91%, respectively. The recovery for all samples was >80%, except for liver samples (>70%). The main pharmacokinetic parameters obtained were: T(max) = 0.65 +/- 0.07 h, C(max) = 1.5 +/- 0.2 microg/mL, AUC = 34 +/- 6 h microg/mL and K(a) = 3.5 +/- 0.6/h. The experimental results showed that ligustilide was easily absorbed, but its elimination was slow, from 3 to 12 h after oral administration. The concentrations of ligustilide in rat cerebellum, cerebrum, spleen and kidney were higher than those in other organs.     

A rapid urine creatinine assay by capillary zone electrophoresis.

Using capillary zone electrophoresis, the urine creatinine (uCr) assay was validated in extemporaneous diluted urine, both in healthy subjects and athletes, with the uCr concentration as a reference value to compare excretion rates of other metabolites in the same samples. The electrokinetic sample injection was carried out at 10 kV per 10 s; UV absorbance detection was at 254 nm. Using standard samples, the creatinine migration mean time in 100 mmol/L acetate buffer, pH 4.4, was 3.3+/-0.2 min; the repeatability for absolute migration mean time was 0.6% and peak height repeatability was 2.9%. The correlation coefficient of the standard curve was r = 0.999 and the detection limit was 23.1 micromol/L. Intra- and interassay coefficients of variation (CV) were 3.0 and 3.6%, respectively; recovery was 99+/-3% and linearity was r= 0.98. Normal urine samples were diluted 1:80 in run buffer. The present CE urine creatinine assay showed a good correlation with HPLC and with Jaffe methods (r = 0.98 and r = 0.97, respectively; p < 0.0001). The uCr in the morning urine samples of 34 healthy males (M), 38 healthy females (F), and 83 male athletes (A) was 10.4+/-6.1 mmol/L, 10.8+/-8.1 mmol/L and 13.2+/-6.5 mmol/L, respectively. The uCr difference (p < 0.02) between M and A and a correlation (p < 0.05) with age in A were observed.     

The interaction at the solid/liquid interface of 2,4-dichlorophenoxyacetic acid with silica modified by reaction with ammonia gas

A method for the attachment of ammonia to modified silica gel has been developed. In the first step, tetraethylorthosilicate and (3-glycidoxypropyl)trimethoxysilane were co-condensed in the presence of n-dodecylamine, a neutral surfactant template, to produce a modified mesoporous silica. The epoxy group incorporated into the mesoporous silica was opened by ammonia gas thus introducing amine chelate groups covalently bound to the inorganic surface. The modified material contained 1.13+/-0.06 mmol of amine per gram of silica, exhibited a surface area of 831+/-29 m(2)g(-1) and a porous diameter of 1.95 nm. Infrared, 29Si and 13C NMR spectra were in agreement with the proposed structure of the modified mesoporous silica in the solid state. This ordered organic-inorganic hybrid presented a high capacity for the removal of the agrochemical 2,4-dichlorophenoxyacetic acid from water: the DeltaH and DeltaG values for the interaction were determined to be -110.61 and -9.37 kJ mol(-1), respectively.