Integrated approach to surfactant environmental safety assessment

J. R. Simms, D. A. Woods, D. R. Walley, T. Keough, B. S. Schwab, and R. J. Larson. Anal. Chem. , 1992, 64 (23), pp 2951–2957. DOI: 10.1021/ac00047a0...
0 downloads 0 Views 872KB Size
Anal. Chem. 1992, 64, 2951-2957

2911

Integrated Approach to Surfactant Environmental Safety Assessment: Fast Atom Bombardment Mass Spectrometry and Liquid Scintillation Counting To Determine the Mechanism and Kinetics of Surfactant Biodegradation J. R. Simms, D. A. Woods, D. R. Walley, and T. Keough' The Procter and Gamble Company, Miami Valley Laboratories, P.O.Box 398707,Cincinnati, Ohio 45239-8707

B. 5. Schwab and R. J. Larson The Procter and Gamble Company, Ivorydale Technical Center, Cincinnati, Ohio 45217

Fast atom bombardment ma- spectrometry and liquid scintillation countlng have been used to study the biodegradation of a novel cationk surfactant in live sludge. The rates of prlmary biodegradation and the extent of complete mlneralzatbnwere detennhed. Fwthermore,an intermediate degradation product was identlfled and Its rates of formation and subsequent removal have been establlshod. These data find utility in assessing the envlronnental safety of the surfactant and the accuracy of various envlronmentai fate models.

INTRODUCTION Annual worldwide production of surfactants currently exceeds lo9 pounds.' These materials are widely used in detergent produde and have the potential for broad-scale release into aquatic and terrestrial environments. Biodegradation, an important removal mechanism, plays a major role in minimizing Surfactant environmental i m p a ~ t .Ac~~~ cordingly, analytical methods capable of accurately assessing the extents and rates of surfactant biodegradation have been developed. Methods to follow both primary biodegradation, i.e. the initial degradation or alteration of the parent surfactant, or complete mineralization, i.e. conversion to COZ, water, and inorganic salts, have been reviewed by Swisher.' They include nonspecific metabolic methods, titrations, and chromatographic techniques such as desulfonation GC, TLC, and HPLC. Desorption ionization (DI) mass spectrometry is a powerful set of tools for the characterization of nonvolatile compounds including organic ~urfactants."'~ In the context of environ(1) Llenado, R. A.; Neubecker, T. A. Anal. Chem. 1983,55,93R-l02R. (2) Larson, R. J.; Bishop, W. E. Soap, Cosmet. Chem. Spec. 1988,64, 58-122. .. (3) Larson, R. J.; Bishop, W. E. HAPPI, Household Pers. R o d . I d . 1988,25,84-98. (4) Swisher, R. D. Surfactant Biodegradation; Marcel Dekker: New York, 1987. (5) Burlingame,A.L.;Whitney,J. 0.;Russell,D. H.Anal. Chem. 1984, 56,417R-467R. (6) Burlingame, A. L.; Baillie, T. A.; Derrick, P. J. Anal. Chem. 1986, 58,165R-211R. (7) Pachuta, S. J.; Cooks, R. G. Chem. Reu. 1987,87,647-669. (8) Busch, K. L. In Methods of Surface Characterization;Czanderna, A. W., Hercules, D. D., Eds.; Plenum Surface Science Series; Plenum: New York, Vol. 3 (in press). (9) Soft Ionization Biological Mass Spectrometry;Morris, H. R., Ed.; Heyden: London, 1980. (10) Desorption Mass Spectrometry; Lyon, P. A., Ed.; American Chemical Society Sympoeium Series No. 291; American Chemical Society: Washington,DC, 1985. ~~~

0003-2700/92/0364-295 1503.00/0

mental analyses, field desorption mass spectrometry has been previously used to identify and quantitate trace levels of nonionic surfactants in river waterls and to follow the biodegradationkinetics of a cationic and a nonionic surfactant in a river die-away test.16J7 We have recently reviewed quantitative applications of another desorption ionization method, fast atom bombardment mass spectrometry (FAB/ MS), and demonstrated that FABIMS could be wed to accurately quantitate low part-per-billion levels of cationic surfactants in aqueous environmental matrices.18 The principal advantages of the DI mans spectrometry methods, relative to other analytical approaches, are sensitivity, selectivity, and speed of analysis. In this paper, we report a new analytical approach, based on FABIMS and liquid scintillation counting, to determine surfactant biodegradation at the molecular level. We have used the method to monitor the biodegradation of low partper-milliom (ppm) levels of a new cationic surfactant in a complex environmental matrix, i.e. sludge. We have determined the rate of surfactant primary biodegradation, identified an intermediate degradation product, determined its rates of formation and removal, and established the extent of complete mineralization to COZ. These data, coupled with toxicity results and fate modeling, form the basis for the environmental safety assessment of the parent surfactant.

EXPERIMENTAL SECTION Synthesisof Surfactants. N-Octadecyl-N-[@almitoylozy)ethyl]-N,N-dimethylammonium chloride (III) was prepared according to Scheme I. 2-(Methylamino)ethanol(AldrichChemical Co., Milwaukee, WI) was alkylated with 1-bromooctadecane (AldrichChemical Co.) in refluxing 1,2-dmethoxyethane in 93 % yield. The amine (I) was esterified with palmitoyl chloride (Aldrich Chemical Co.) in the presence of TEA (triethylamine). The resulting amine ester (11)was washed with 1,2-dimethoxyethane and fiitered to remove inaolubles. It was then rewashed with acetone. Structure elucidation was carried out with 'H and '*C NMR, IR, and TLC (51 CHCWCHsOH). The ester amine (11) Cochan, R. L. Appl. Spectrosc. Rev. 1986,22, 137-187. (12) Turner, N. H.; Dunlap, B. I.; Colton, R. J. Anal. Chem. 1984,56, 373R-416R. (13) Sundqvist, B.; Macfmlane, R. D. Mass Spectrom. Rev. 1986,4, 421-460. (14) Wood,G. W. Mass Spectrom. Reu. 1982,1,63-102. (15) Shiraiehi, H.; Otauki, A.; Fuwa, K. Bull. Chem. SOC.Jpn. 1982, 55.1410-1415. '(16) Scheider,E.; Levsen, K. Comm.Eur. Communities, [REP] EUR 1986, NO.EUR 10388,14-25. (17) Scheider, E.; Levsen, K. Fresenius' 2.Anal. Chem. 1987,326, 43-48. ~(18) Simms, J. R.; Keough, T.;Ward, S. R.; Moore, B. L.; Bandurraga, M. M. Anal. Chem. 1988,60, 2613-2620. 0 1992 Americen Chemkal Society

2952

ANALYTICAL CHEMISTRY, VOL. 64, NO. 23, DECEMBER 1, 1992

Scheme I H

0

~

-hi

HOCHpCHp-N-CHs --)HOCH,CH,-N-CH,

I

I/

C15COCHpCH,-NCH3 Primary Biodegradation

Complete Mineralizahon

99%

i

dose t L sludge volumes I 50 ml timed aliquols add '3C mte1naI standard(s) equilibrate,quench

1%

I

Radiocount Total dosed radioactivity=A

I

,yophilire

I

extract 3x with warm EtOH EtOH Insoluble I residue and glass wool filter

(11) was then quaternized with methyl chloride (J. T. Baker I add 1 ml H 0 and 10 mls So?uene Chemical Co., Phillipsburg, NJ) under atmospheric pressure in radiocount I an inert atmosphere (argon)and the reaction was monitored with heat at 42" for 24 hours concentrate *g lo dryness (EtOH soluble TLC (silica gel, CHCl&HsOH 5:l). The resulting product (111) radioactivity) add 5 mls EtOH prep TLC I was suspended in l,2-dimethoxyethaneI filtered, and dried to yield 99% pure quaternary salt (purity verified by quantitative two-dimensional (2D) TLC and structure by 'H and 13CNMR, IR, and FAB/MS). N-(2-Hydroxyethyl)-NJV-dimethyloctadecylummnium chloride (ZV)was prepared from I by reaction with methyl chloride at room temperature and atmospheric pressure (under argon). Total recovered radioactivity= B+C+D Completion of the reaction was determined by TLC (silica gel, CHCUCI&OH 51). The product was recrystallizedwith hexanes and 2-propanol,filtered, and dried. Purity (81% ) was determined by quantitative 2D TLC, and structure verification, by lH and 13C NMR, IR, and FAB/MS. Evolved COS. The method for direct determination of evolved ~ ~ w* i~l l only be briefly WOZhas been described p r e v i ~ u s l yand Synthesis of Stable Isotope Standards. [lSC]-N-Octadecyldiscussed here. The equipment consisted of 2-L Erlenmeyer N-[ @almitoyloxy)ethyl]-N,N-dimethylammonium iodide was flasks containing 1L of activated sludge. The inlet porta of the prepared by reaction of 1.2 g of I1with W H d (0.45 g, Cambridge flasks were connected to a COz free air supply with Tygon tubing. Isotope Labs, Woburn, MA) in absolute CzHsOH at 85-90 "C for The exit porta were connected with Tygontubing to three French 18 h to yield 1.5 g of quatemized amine. The product was squares (in series) each containing 100 mL of 1.5 N KOH. The recrystallized using 101:0.08 hexanes:2-propanol:CHCl3 to give inlet air flow,regulated at 50 mL/min, was used to sweep evolved 89% purity and its structure was determined using TLC, lH and l4COZinto these external KOH traps for later analyses. 13C NMR, IR, and FAB/MS. The Erlenmeyer flasks were dosed with 5 mg of [14C]-IIIand [~~C]-N-(2-Hydroxyethyl)-N,N-dimethyloctadecylammoni-incubated with shaking (100-150rpm)for 30 min. Then triplicate um iodide was prepared by reaction of 1.046 g of I with 13CH31 2-mL samples were syringe collected for radiocounting. These (0.68 g) in refluxing absolute CzHsOH. The reaction was values represent the total dosed radioactivity and were used to monitored for completion by TLC. Quaternized amine (1.08 g) normalizeall subsequent measurements as a function of time. At was obtained and was recrystallized in hexanes:2-propanol(20 various time intervals, 1-mL samples were withdrawn from the 1). Purity (80%)was assessed by quantitative 2D TLC, and the external KOH traps and radiocounted. At the same time the structure was verified by 'H and l3C NMR, IR, and FAB/MS. external KOH traps were sampled, 6 mL of the aqueous sludge was withdrawn from the Erlenmeyer flasks and added to serum Synthesis of Radioactive Isotope Standards. [I'CI-Nvials with reservoire housing fluted paper wicks that had been Octadecyl-N-[ @almitoyloxy)ethyl]-NJV-dimethylammonium saturated with 200 pL of 1.5 N KOH. The sealed serum vials chloride (ZZI) was prepared by reacting 91.5 mg of I1with 14CH31 were acidified with 200 pL of concentrated HC1 to release (2.57 mg, Amersham U.K.) and CH31(20.6 mg, Aldrich Chemical Hz1'CO3and/or Hl4CO3trapped in the aqueouslayer. These vials Co.) in a reactivial for 48h at 80 OC. The reaction was monitored were incubated at room temperature for 24 h. After 24 h, the by TLC. The product was dissolved in 7030 CzHsOH/HzOand paper wicks were radiocounted. The total evolved WOz was placed on an ion-exchange column (Dowex-1, C1- resin). The then calculated as the s u m of that collected in the extemal base isolated product was shown to have 89% radioactive purity with traps and that collected on the paper wicks (after appropriate a specific activity of 9.5 mCi/g by radioassay. volume corrections). All valueswere then normalizedto the initial [14C]-N-(2-Hydroxyethyl)-N,N-dimethyloctadecylammoni- dose. um bromide (IV)was prepared by reaction of 155 mg of I with Batch-Activated Sludge System. Livesludge was obtained 14CHd(17.7mg) in a reactivial in refluxing toluene/acetonitrile. from the Colerain Heights Sewage Treatment Facility in CinThe reaction was monitored by TLC. The product was passed cinnati, OH. The sludgewas come filtered and suspended solids through an ion-exchangecolumn (Dowex-1,C1-resin) and verified were determinedz1 (1950-2200 mg/L) prior to use. One liter by TLC (silica gel, CHCl&HsOH/HzO/formic acid 80:26:3:1). portions of near-homogeneoussludge were measured into gradRadioassay showed >99% radiopurity and the specific activity uated cylinders and transferred into 2-L Erlenmeyer flasks.The to be 29.1 mCi/g. flasks were continuously aerated and gently shaken on an AquaTherm water bath shaker (New BrunswickScientific, New General Procedure. The general method to follow surfactant Brunewick, NJ) until s t a r t of the biodegradation experiment biodegradation in sludge is summarized in Figure 1. Primary (generally within 1 day after receipt). Deactivated sludge was biodegradationwas determined at the molecular levelusingFAB/ MS as outlined on the left-hand side of the figure. The extent (19)Larson, R.J.; Payne, A. G. Appl. Environ. Microbiol. 1981,41, of completemineralization of the surfactant was estimated using 621-627. the radiocounting experiments outlined on the right-hand side (20)Larson,R. J.;Games,L. M.Environ.Sci. Technol. 1981,15,1488of the figure. These latter estimates were compared to exper1493. imentally determined levels of evolved l4COZ,values that were (21)Hobbie, J. E.;Daley, R. J.; Jasper, S. Appl. Environ. Microbiol. 1977,33,1225-1228. directly measuredusingthe independent method discussed below.

F

pzzZ-1

ANALYTICAL CHEMISTRY, VOL. 64, NO. 23, DECEMBER 1, 1992

used for some of the control studies and in the estimation of the extent of completemineralization of ID. For those studies, sludge was deactivated either by lowering the pH to about 3 or by addition of aqueous formaldehyde (3%). Equivalent FAB/MS and radiocounting results were obtained from sludge deactivated by either method. Dosing Solutions and Internal Standards. The live and deactivated sludge flasks were dosed with the “cold”surfactant I11and its corresponding 14Canalog. The dosing solutions were freshly prepared before each study in the followingproportions: 4 mg of I11 and 1mg of its IW analog (corresponding to about 20 X 106disintegrations per minute (dpm)) per 2 mL of absolute CzHsOH. Two milliliter volumes of this solution were used to dose each flask. A few microliters of the dosing solutions were analyzed by FAB/MS with full scanning and multichannel signal averaging, over the molecular ion region, to ensure integrity just prior to use. Furthermore, about 1%aliquots of the solutions were radiocounted (A, Figure 1) to establish exactly the total radioactivity dosed into each flask at the start of the study. Separate solutions of the 13Canalogswere prepared as internal standards for the quantitative FAB/MS studies. In our initial experiments, we quantitated only removal of 111. A freshly prepared CzHsOH solution of its 13C analog, having a concentration of 200 ng/pL, was used. In subsequent studies, we quantitated both the removal of (111)and the formation/removal of its initial hydrolysis product (IV). For these latter studies, internal standard solutions containing both of the ISCanalogs of I11and IV, with concentrations of 200 and 50 ng/pL, respectively, were used. Primary Biodegradation. Both live and deactivated sludge flasks were dosed with 2 mL of the lzC/lW dosing solution, thoroughly mixed, Stoppered, and placed onto the shaker. The flasks were sampled at predetermined time points such as 1,3, 5,7,9,10.5,12,24,36,48,72, and 100 h by withdrawing 50-mL aliquots into volumetric pipets using an electric pipet-aid (Drummond Scientific Co., Broomall,PA) that greatly facilitated rapid sampling. Each flask was then aerated for a few minutes, stoppered, and returned to the shaker until the next sampling time. Thisparticular aeration procedure was employedbecause, in preliminary experiments, we found continuous aeration was unnecessary and it led to excessive evaporation over the time course of the measurements. Each of the 50-mL timed aliquots was immediately dosed with 500 pL of the appropriate 13Cinternal standard solution (a total of 100pg of 1% analog of I11and either 0 or 25 pg of the 13Canalog of IV), mixed, and allowed to equilibrate with gentle shaking for 15min. Biodegradation was then quenched by rapidly freezing the sample in a dry ice/acetone bath. The samples were stored in a freezer at 0 OC until lyophilized overnight in a Cons01 12 lyophilizer (Virtis Co., Inc., Gardiner, NY). The residues from the lyophilizedsamples were then extracted with 50 mL of warm CzHsOH for 2 min and filtered over silanized glass wool into 250-mL round bottom flasks. The extraction/ filtration step was repeated two more times for each sample so the final volumes of the extracts were 150 mL. The CzHsOH insoluble particulates and glass wool filters were recovered, digested by the method discussed below, and radiocounted (B and C, Figure 1). Furthermore, 1.5 mL of each of the CzHsOH extracts (1%)was saved for radiocounting (D, Figure 1). The remaining 99% portions of the extracts were concentrated to dryness on a rotoevaporator and reconstituted into minimum volumes of CHCb (