Determination of dimethylethylamine in air samples from iron foundries

C. Determination of dimethylethylamine in air samples f1"om iron foundries by isotachophoresis. Scand J Work Environ Health II (1985) 307-310. A method for the determination of dimethylethylamine in workroom air has been developed. Static gas standards of the amine in air were used to evaluate different sampling techniques. The analysis was performed by isotachophoresis. The sampling equipment of choice was midget impinger flasks of glass or polystyrene containing 10 ml of hydrochloric acid (50 mmolll). The method was used in a field study of three dif ferent iron foundrie s where air samples were taken. The dimethylethylaminc concentration found was in the range 0.5-155 mg/m'. No pretreatment of the samples was necessary, and no interfering substances from the air in the foundries affected the analysis.

In the production of mold cores in the foundry industry the cold-box techn ique is of increas ing importance. The mold cores are made of fine granular sand, 4,4'-diphenylmethane diisocyanate (MDI), and a liquid phenolic resin. The reaction is catalyzed by a gaseous tertiary amine which is pres sed through the mixture. Triethylamine (TEA) has thus far been the dominating amine catalyst. But for the last few years the use of the more volatile dimethylethylamine (DMEA) seems to be increasing in Sweden .
DMEA is a liquid with a strong ammoniacal odor. It has a boiling point of 35°C and is very flammable with a flashpoint of -36°C (8). In animal experiments it has been indicated that DMEA is very irritating to the eyes, remarkably irritating to the mucous membrane in the respiratory organ, and slightly irritant to the skin (8). The hygienic standard for DMEA in workroom air is 75 mg/m' in Denmark, whereas no corresponding value has been adopted in Sweden .
Gas chromatography is generally used for the analysis of DMEA. Two such methods using aromatic polymers or ethylene glycols as stationary phases have been described by Lindsay Smith and co-workers (6, 7) for the analysi s of DMEA and some other aliphatic amine s. Casselman & Bannard (I) have indicated the problems connected with the gas chromatographic analysis of amines. Recently Dalene et al (2) published a method for the analysis of TEA and oth er aliphatic amines by gas chromatography. Volpi (10)  method to measure DMEA and several other compounds in exhaust gases from different proce sses in foundries, among them cold-box core-making.
In connection with an extensive study on the occupational environment in foundries and as part of our investigations on anal ytical methods for amine s used in indu str y (3,4, 5), we have developed an alternative method for the analysis of DMEA in air samples by isotachophoresis (ITP). The method has been used in a field stud y of three different foundries and is presented in this report.

Preparation of gaseous standards
Static gaseous standards of DMEA in air were made in home-made sacks, as shown in figure I. A sheet of aluminum laminated with a 12-Jtm thick polyester layer (Akerlund & Rausing, Lund, Sweden) was folded to form a sack (125 x 80 em) with the pol yester layer inside. The edges were affixed with tape adhesive on both sides. The inlet system for the diluting air was a polytetrafluoroethylene (PTFE) tube (inner diameter 4 mm) fastened with PTFE nuts directly on the wall of the sack. Purified diluting air (charcoal filter) was introduced into the sack, and the volume (generally 100 I) was registered on a gas meter. DMEA was injected directly through the laminate by a microliter syringe. In this step the DMEA and the syringe had to be chilled to avoid losses due to evaporation. A gentle tapping of the sack was used to mix the gas. Conc entrations between 15 and 75 mg/m' were generated . of hydrochloric acid (HCI) (50 rnmol/I) was compared by bubbling the gaseous amine standard through home-made midget impinger flasks of polystyrene, spill-proof midget impinger flasks of glass (Svenska labglas AB, Stockholm, Sweden), and gas washing bottles of glass. Sampling was also performed on silica gel tubes [6 mm outer diameter X 70 mm, no 226-10, SKC Inc, Eighty Four, Pennsylvania, United States (USA»). When comparisons were made with glass flasks or silica gel tubes, they were run in parallel with the polystyrene flasks . Portable sampler pumps (Anatole J Sipin Co, New York, New York, USA) were used at a flow rate of 200 ml/min. During the sampling the flow rate was intermittently checked with a soap film meter connected to the outlet of the pump.
For desorption of the amine from the silica gel the following solvents were used: water, HCI (50 mmol/I), and HCI (100 mmol/I) : ethanol (1 : 1). Each gel section was transferred to separate glass vials, 1.0 ml of the desorption solvent was added to the vial with the sorbent layer, and 0.5 ml of the the same solvent was added to the back-up so rbent layer vial. The vials were then shaken for 30 min in a "laboratory shaker.
Foundry study. Air samples were taken in three different iron foundries. Foundry A had several small and rather old machines in the area where small mold cores « I kg) were produced. Five persons were exposed. Foundry B had two modern machines which were fully enclosed. Mold cores of small and medium sizes (1-4 kg) were produced, and four persons were exposed . Foundry C had one modern large machine (enclosed) , and one old machine in which large and medium-sized mold cores (5-15 kg) were produced. Three persons were exposed. In all three foundries the core machines were ventilated, and in foundries B and C the exhaust gases were burned in a butane flame .
Both personal and stationary samples were collected. The stationary samples were taken about 1.5 m above the floor level close to the machines and in the area where newly made cores were stored. At all sampling sites polystyrene flasks were used. When the samples were collected in glass flasks or on silica gel tubes, they were run in parallel with the polystyrene flasks. The sampling occurred on 2 d when the production was said to be normal.

Concentration of the samples
The sensitivity of the analysis can be increased by concentration of the absorption solution from the impinger flasks. The procedure involves transferring 5 ml of the solution to a polypropylene tube, which was put in a thermostated heating block, and evaporating the solution to dryness at 80 cC. The process was speeded up by passing a gentle stream of nitrogen over the surface of the solution. To avoid losses, probably due to sublimation of the hydrochloride of the amine, the process was stopped when dryness was reached in the tubes. The evaporation rate for the absorption solution was 1 ml/h . The residue was dissolved in 500 /ll of distilled water. An aliquot of the solution was injected directly into the isotachophoretic instrument for analysis .

Analysis
The analysis was performed by isotachophoresis with an LKB 2127 Tachophor (LKB-Produkter AB, Bromma, Sweden) equipped with a conductivity detector and a capillary tube (PTFE, inner diameter 0.5 mm, length 230 mm). The capillary was thermostated at 15 cC. The leading electrolyte was potassium hydroxide (10 mmol/I) in 0.4 0J0 (weight/volume) hydroxypropylmethylcellulose (90 HG 15 Pas, Dow Chemicals, Midland, Michigan, USA). The pH of the solution was adjusted to 8.8 by the addition of L-Valine (solid) (Sigma, St Louis, Missouri, USA). The terminating electrolyte was tris(hydroxymethyl)aminomethane (20 mmol/I) in HCI (5 mmol/l), pH 8.5. The sample volume was between 2 and 10 /ll, the migration current was 100 /lA, and the time of analysis was about 12 min. DMEA was quantified from measurements of the zone lengths on the graph through a pocket lens, magnifying 10 times, with a graduated scale.

Results and discussion
Laboratory experiments An analytical separation of DMEA and TEA is shown in figure 2. A calibration graph for DMEA analyzed by isotachophoresis was made from standard solutions of DMEA in water. It was linear in the range tested (I to 150 nmol). The practical detection Consequently the worker who take s the mold cores out of the machine will ha ve a higher exposure than the stationary samples indicate.
AIl samples in the field study were taken at random, and the investigation is not to be regarded as an occupational examination of the exposure to DMEA in 1 min I I limit for DMEA in this system was about I nmol.
(This value corresponds to 7.3 /Lmol of amine in a sample concentrated 10 times, which can be obtained from 4 I of air containing a DMEA content of 2 mg/m' .) The result s of the laboratory experiments of sampling DMEA from a sack with two dif ferent impinger flasks are shown in table I . By "overaIl recovery" we mean the quo tient of the DMEA concentra tion determined by the isota chophoretic anal ysis to the calculated value of the concentration in th e sack. The uncertain ty of the determin ations includes error s from the preparation of the gas standard, sampling, concentration of the absorption solution, and the final analysis. The recoveries were similar for the two types of impinger flasks.
If necessary the absorption solution can be concentrated 10 to 20 times with negligible loss of DMEA.
In one experiment gas washing bottles of glass with sinter ed glass filters were used for sampling about 400 /Lg of DMEA from gas stand ard s. The recoveries had a slightly greater variation in comparison to those obtained when impinger flasks were used (recovery = 96 0/ 0, SD = 13, N = 5).
Th e experiments with adsorption tube s gave unsatisfying recoveries with the desorption solutions used. Ho wever, the use of solid sor bents for the sampling of amines wiII be dealt with in anot her work .

Foundry study
The results of the determinat ions of DMEA in three different foundries are shown in table 2. The figures indicate that the personal samples are slightly higher than the stationary ones tak en in the vicinity of the Time Figure 2. Isotachophoret ic separ ation of dlmethylethylamine (23 nmol, 1) and tr iethylamine (20 nmol, 2). The migration cu rrent was 50 p.A, and the scale ind icates 1 min of recording.
(L = leading ele ct ro lyte, T = term inating el ectrolyte , 0 = dif· fe rent ial signal) Table 1. Analys is of dim ethy lethylamine (OMEA) from static gas standards in an aluminum·lam inated pol yester sac k. Comparison bet ween sampl ing in midget im pin ger flas ks of pol ystyrene and glass at thr ee d ifferent concentration levels.  DMEA in air at a concentration of 2 mg /m' in iron foundries w hen sa m p ling 4 I of air. No pretreatment of the samples before analysis is necessary, and no interference affects the analysis. The unique quality of the isotachophoretic analysis to discriminate interfer ing substa nces w hic h are not charged at the prevailing pH o f the system makes the evaluation of the separation rational and convenient.