Parfait column

The remainder of the modified parfait column consisted of an MSC-1 cation-exchange and an A-162 anion-exchange bed. The elution conditions for these beds were modified to minimize contamination of eluates and to selectively desorb organic anions and cations. With the modified protocol, 20 model compounds (Table I), selected by the U.S. Environmental Protection Agency (USEPA) Health Effects Research Laboratory (HERL), were used to evaluate the recovery efficiency of the method. Recoveries were determined in the presence of 2 ppm of a humic acid supplied by HERL. 

Instrumentation. Vacuum distillation of parfait column effluents was performed on an FTS Systems model FD-20-84, high-capacity, corrosion-resistant, freeze-drying apparatus modified as described in the text. 

Basic Procedures. The synthetic hard water used in all of the parfait column studies consisted of 0.070 g of NaHCC>3, 0.156 g of CaS04 2H20, and 0.047 g of CaCl2 per liter. This solution was prepared in 8-L batches just before use. The final pH was 7.2. 

For parfait column studies, standard solutions (400 /ug/mL) of each organic solute except phenanthrene were prepared in a suitable solvent (acetone, water, or methanol), and 1.00 mL of the standard was added to 8 L of the synthetic hard water with rapid stirring. This yielded a final organic solute concentration of 50 ng/L and a solvent concentration of less than 126 mg/L. In the runs simultaneously involving 16 compound four different mixtures of solutes were prepared so that the final solvent concentration was 500 mg/L. Phenanthrene was spiked to a final concentration of 1.0 /zg/L, rather than 50 /zg/L. Solutions were not degassed before application to a parfait column. No adjustments of pH were necessary after adding solutes to the synthetic hard water. 

The Teflon and ion-exchange beds were eluted with the eluents as follows The addition funnels containing each bed were separated from the parfait column, and a separate, empty, addition funnel was loosely mounted on top to supply eluents. The bottom stopcock was closed, and the first eluent was slowly added until air escaped the bed and the bed was covered by 0.5 cm of liquid. The upper addition funnel was then firmly seated, and elution began by opening the bottom stopcock. Flow was adjusted to less than 3 mL/min with... 

The aqueous effluent from the parfait column was exposed to vacuum in a silanized 12-L flask on a high-capacity freeze dryer. The effluent was stirred with a magnetically coupled stirring bar and heated by immersion of the lower half of the flask in a 30 °C constant-temperature water bath. The rate of distillation was controlled to about 8 L/24 h with a needle-valve air bleed. Distillation was continued to a volume of about 500 mL. Then the concentrate was transferred to a smaller flask, and the volume was further reduced to 10 mL on a conventional rotary evaporator. 

Reconstruction experiments showed that methylation efficiency was the same in the solvent used to prepare the standard and that obtained from experimental parfait column concentrates. Chromatography was usually completed within 36 h of final concentration or methylation of a sample. Samples were stored in Teflon-lined crimp-sealed vials at —20 °C. 

Concentrates of parfait columns were routinely assayed within 48 h of preparation. Slow evaporation of FI was observed in some experiments, even when Supelco Teflon-lined septum vials were used. Individual concentrates were rechromatographed at various times, and it appeared that profiles could not be reliably reproduced after several days, even when the concentrates were stored at —20 °C. Certain compounds, particularly BHT and furfural, were found to change their profiles in a matter of days, probably because of decomposition. 

Influence of Hypochlorite on Parfait Columns. One potential use of the parfait method is the recovery of organic matter from drinking water. To test for the interaction of chlorine disinfectant with column components or eluents, the influence of 2 ppm of hypochlorite was assessed in an unspiked control column. Each eluate was assayed for hypochlorite by using the ferrous N,N-diethyl-p-phenylenediamine titrimetric method (12). No hypochlorite was detected. Each eluate was also analyzed by GC and found to be virtually identical to a blank column without hypochlorite run simultaneously. 

Figure 6. Gas chromatograms of parfait column fractions F3 (top), (middle)y and F5 (bottom) from an unspiked control column.

HUMIC Acid. Humic acid did not contribute detectable impurities to the eluates of blank parfait columns. This result was apparently due to the insolubility of humate in the organic solvents used to elute the Teflon and ion-exchange beds and the inability of the humate to volatilize in the GC. Humic acid did, however, distribute itself throughout the parfait column, as indicated by the observation of color entering the column effluent, F7. When 16 mg of humate in 8 L of synthetic hard water was passed through a parfait column having the Teflon bed divided into three sequential 50-mL beds, 8.9 , 5.0 , and 2.9 of the total humate were found in the aqueous phases that separated upon elution of these beds, as indicated by absorbance at 200 nm. The column effluent from this experiment contained 5.1 of the humate applied. The majority of the humate applied was found as color adsorbed to PTFE, and it did not elute into methylene chloride. Conditions to elute it from PTFE were not explored. 

Table VIII. Recovery of Glycine on Parfait Column...

Further elution of MSC-1 after standard elution yielded fraction F2. e One-tenth scale standard parfait column using 14C-glycine. Eluates and drying-conditioning washes from individual ion-exchange resins were pooled Teflon, MSC-1, and A-162 denote radioactivity on these beds after elution (see text). 

In column 110, it is also theoretically possible that glycine com-plexed with the added humic acid and that it was sequestered in the aqueous phase of the Teflon eluate and bound to the Teflon bed. To test this explanation, a 1/10-scale parfait column was constructed and 4 liCi of 14C-glycine, 40 fig total, was applied in 800 mL of synthetic hard water (column 123). In this experiment, the alcohol and solvent 1 conditioning washes were combined with the standard eluates of each bed before counting. These solutions were not concentrated before counting. Quench correction was by the channels ratio method. 

Recovery of Other Solutes on Standard Parfait Columns. The porous Teflon bed gave the most reproducible recoveries of test solutes. Table VI identifies the solutes found exclusively on Teflon and shows their recoveries. The influence of increasing the Teflon bed volume from 50 to 150 mL was mentioned earlier. From column 106, it is evident that BHT and stearic acid were the two most strongly adsorbed solutes. The least well-recovered solutes in this group were 1-chloro-dodecane and methyl isobutyl ketone however, both of these solutes are so readily volatilized that losses during concentration of eluates must be considered a likely source of their low recovery. 

Three compounds recovered from parfait columns were also previously tested for breakthrough from 5-mL Teflon beds (6). The capacity factors for these compounds and their recoveries from the Teflon bed of a parfait column showed a rough correlation. Phenanthrene, which was tested in the parfait column only in the presence of humate, was recovered essentially quantitatively from the 5-mL Teflon column and had a capacity factor of 368. About 15 of the caffeine applied to a parfait column in the absence of humate could be recovered from Teflon, and caffeine showed a capacity factor of 22. Only about 2 of the 2,4-dichlorophenol applied to parfait columns could be recovered on Teflon its capacity factor was 5.6. It may therefore be anticipated that compounds following the inverse correlation of solubility with capacity factor and having a capacity factor greater than about 20 should be detectably absorbed to the Teflon bed of a parfait column. Simply increasing the volume of the Teflon bed may also increase the absolute recovery of adsorbable solutes that have modest values of kFor this reason, a 150-mL bed of Teflon per 8 L of water may not be the ideal bed size a larger bed may be better. 

Unrecovered Solutes. Although not all of the applied solutes were recovered in parfait column eluates, reasonable suggestions can be made about the locations of the missing compounds. For example, the unrecovered methyl isobutyl ketone, 1-chlorododecane, and chlorobiphenyls were surely lost from the FI eluate by vaporization during concentration. 

---