m INVESTIGATION OF OAS CHROMATOGRAPHIC

SEPARATION OF TASTES AND ODORS

PRODUCED m ACTINOMTCETES

APPROVED\

SSLm v W lJ t, V V p , Professor \j gv*7rw»B»w \j

imh Minor PrQftf&or

QVsH>SAir\:a)> Direcj^r of the Depa^ment of Biology

sfl . I Dean of the Graduate School

AH INVESTIGATION OF GAS CHROMATOGRAPHIC

SEPARATION OF TASTES AND ODORS

PRODUCED BT ACTINOMTCETES

THESIS

Presented to the Graduate Council of the

North Texas State University in Partial

Fulfillment ©f the Requirements

For th# Degree of

MkStM OF SCIENCE

By

James K. {fetlock, B. A.

Denton, Texas

January, 1965

TABLE OF CONTENTS

Page

LIST OF TABLES iv

LIST OF ILLUSTRATIONS . . . . . . . v

Chapter

I. INTRODUCTION . . . . . . . 1

II. METHODS B

III. RESULTS . . . . . . . . . . . . 19

IV. DISCUSSION . 27

SUMMARY AND CONCLUSIONS 32

BIBLIOGRAPHY 33

ill

LIST OF TABLES

Table Pag© 1. Retention Haas of Three Concentrates

on Carbowax 20-M at $5° Centigrade, laliua at 100 cc. Per Miimt® 20

II. Retention Ti®es ©f Three Concentrates» Toflon Golwaa at 50® Centigrade, fteiiua at 25 ce. For Minuto . . . . . . . . . . 23

iv

LIST OF ILLUSTRATIONS

Figure page

1. Concentrate from "Earthy" Column on • Carbowax 20-M 22

2. Concentrate fro® "Earthy" Pond on Carbowax 20-M 22

3. Concentrate from Blank Column on

Carbowax 20-M . . . . . . . 22

4. Concentrate from "Earthy1* Column on Teflon . . . . 25

5. Concentrate from "Earthy" Pond on Teflon 25

6. Concentrate from Blank Column on Teflon 25

CHAPTER I

INTRODUCTION

The problem of obtaining a supply of fresh water that

is free not only from harmful organisms and chemicals, but

also from unpleasant tastes and ©dors, ha® plagued mankind

throughout it® history. In some cases, the latter criterion

has even been placed above that of the sanitary quality.

Pipe® (9) has cited a ca#e in which people preferred a

cholera-contaminated water to an uncontaainated source which

had an unpleasant taste*

A great deal of progress has been made toward deter-

mining the bacteriological and chemical quality of water

supplies* Many studies have also been done in the area of

tuste and ©dor research, especially in classifying, deter-

mining the strength and cause®, and removing or reducing

tastes and odors.

In 1946, Lendall (6) grouped the cause® of tastes and

odors in surface waters into three large categoriesj

(a) organisms living in the water and their decomposition

products, (b) pollution with domestic sewage or industrial

wastes, and {c\ gases dissolved in the water. Of these

three, the first group, those caused by living organisms, is

the least understood and the hardest to control.

Rohlich and Sarles (10) attempted to relate specific

types of odors in fresh water to a group of aquatic plants,

the algae. It was postulated, but not proved directly, that

"essential oils" producec by some algae and released into the

water may be responsible for certain odors, especially those

described as putrefactive and fishy•

Laughlin (5) described some odors existing in surface

waters as musty, earthy, weedy, woody, grassy, swampy, sour,

fishy, moldy, straw-like, barnyard, potato bin, and septic.

The cause of these odors, however, was not determined.

Silvey £$ &1. (12) conclusively demonstrated that many

tastes and odors occurring in fresh waters, especially

during certain annual periods, could be traced to a group

of fungi, the actinomycetes. They found that some of these

organisms grown in the laboratory upon artificial media

were capable of producing strong odors which were very

similar to those present in many surface waters. Many dif-

ferent strains of these organisms have been Isolated from

bodies of fresh water, and they are also common inhabitants

of soil. Some of the more common odors identified with the

actinomycetes are earthy, woody, musty, marshy, potato bin,

putrefactive, grassy, hay-like, and combinations or varia-

tions of these odors.

Although much has been accomplished toward classifying

the tastes and odors in fresh waters both descriptively and

according to source, little direct evidence has been contrib-

uted toward the chemical identification of the taste and

od©r compounds. It Is known that organic acids are said, to

have sour ©dors and that some amines have fishy odors#

Rohlich and Sarles (10, 11) attempted to relate chemical

structure to ©dor.by comparing odors in waters,to those from

known types of organic compounds. On this basis, it was ,

postulated, that the types of compounds that are most likely

to cause tastes and odors in water are amines, acids, ale®-

hols, esters, heterocyclic nitrogen compounds, and earbonyl

compounds•

Dill (2) and McCernriLck (?) showed that actinooycetes

produce amines, aldehydes, and acids directly as by-products

if* ahnl *1 mm

Pipes (8), using methods of ultraviolet and infrared

absorption spectroscopy of concentrated odor extracts of a

surface water, concluded that the major groups of compounds

present were aromatic amines and aldehydes. The presence of

alcohols, carboxylie acids, esters, and ketones was also

indicated.

Sllvey et al. (13) reported the following compounds

identified from an actinomycete culture: isoamyl amine, iao-

butyl amine, valeric acid, isovaleric acid, beta-hydroxybutyric

acid and isovaleraldehyde,

Prior to the discovery of gas chromatography in 1952 by

Jaaas and Martin (4), chemical analysis of naturally occurring

taste and odor compounds was very difficult because most of

these compounds are present in such aaall quantities. Since

that time arid through its rapid development during the past

decade, gas chromatography is proving to be on® of the most

useful analytical tools in the study of flavors and odors, as

wall a® most chemical compounds, both organic and inorganic*

With the advancement of the so-called ioniiation detectors,

the amounts of chemical compounds which can be detected have

reached such small proportions that new tarns such a# th®

"nanogram8 and th® wpicogramM have been introduced to des-

cribe these quantities {1). One nanogram is 0.001 of a

microgram; one picogram is 0.001 of a nanogram*

In 1963# Gaines and Collins (3) published a significant

paper on the analysis of volatile compounds produced by an

actinomycete. Employing the methods of fractional distilla-

tion, ether extraction, and gas chromatography, they were

able to demonstrate the presence of the fo&owing compounds;

ethanel, ethyl acetate, isobutyl alcohol, isepropyl alcohol,

is©propyl acetate, isobutyl acetate, methyl esters of acetic

acid, fonaic acid and isobutyric acid, and at least two un-

identified compounds. It remains to be determined, however,

if all, or in fact any, of these compounds are responsible

for the odors produced by the organisms*

It would appear desirable from the above discussion to

develop a method by which volatile substances emitted from

solutions could be analysed directly by gas chromatography

without the questionable procedure of previous chemical or

physical treatment* It is the purpose of this investigation

to evaluate a few selected gas chromatographic materials for

the separation of ©dorants produced by an actinomycet® culture

under laboratory conditions. An attempt i» also mad® to com-

pare the results obtained with concentrated solutions of a

surface water containing similar odors under natural conditions.

REFERENCES CITED IN INTRODUCTION

1* 'Aerograph Research Motes, fl©e^l<gals Electron ' Affinity. Wilkens Instrument and Eesearch. Inc«,

' ' d&li/ernla, 1964.

2. Dill, W. S•, "The Chemical Compounds Produced by Actinomycetes and Their Relation to Taste® and Odors in a Water Supply," finpublished master*s thesis, Department of Biology, North Texas State University, Denton, Texas, 1951.

3. (Mines, «. D. and E. P. Collins, "Volatile Substances Produced by Streptomyces odorifer,H Llovdia. XXTI (December, 1963), 247-253#

4. Barnes, A, T. and A. J. Martin, "Gas-kiquid Partition ChroaatOKraphy.* Biochemical Journal, L (June. 1952). 679-690# " '

5. Laughlin, Harold F., ^Palatable Level with the Threshold

XX°(Aug!wt" flllf Journal,

6. Lendall, Harmon N., #A Comprehensive Surrey of the Taste and Odor Problem,n Taste and Odor Control Journal. XII (June, 1946), 1-57?

7. McCoraick, William 6., "The Cultural, Physiological, Morphological and Chemical Characteristics of an Actinomycete from lake Waco, Texas,* unpublished master's thesis, Departaent of Biology, North Texas State University, Denton, Texas, 1954#

$, Pipes, Wesley 0«, Jr., "An Investigation of Naturally Occurring Tastes and Odor® from Fresh Waters,tt

unpublished master's thesis, Departaent of Biolo, j North Texas State University, Denton, Texas, 1951.

9« "An Investigation of naturally dcourring Tastes and Odors from Fresh Waters,'1

unpublished master's thesis, Department of Biology, North Texas State University, Denton, Texas, 1951» citing Jaaes Churchill, Report on the Cholera Outbreak In the Parrish §£ St. Jaaes. lest^OTster.'' during the™ iutuan of 1854. presented to the fesCry by the Cholera inquiring SoSKittee, July, JUI55*

10. Rohllch, Gt©rg@ A* and William B. Sarlss, "Chsmical Corn-

Position of Algae and It® Relationship to Taste and dor," Tasty and Odor Control Journal, XYIII {November, 19E9T, T?7.

11. . nThs Chemistry "" of Organic Compounds Kssponsible ror tastes and Odors,w

fasts and Odor Control Journal. X? (Juno,,1949)» 1-5•

12. Silvay, J. K. 0., James C. Russell, and David R. Eoddsn, "Actinomycetes and Common Tastes and Odors," Journal of American Watsr Works Association. XLII 17«iMry, ^

13* Silrsy, J* K. G., James C. Russell, B. E. Redden, and W. McCormick, "Actinomycetes and Common Tastes and Odors,* Journal of American Water Works Association. m (March, WJfi)7'" """ —

CHAPTER II

METHODS

Several texts are available which give' excellent

explanations of the theory and practice behind gas chro-

matography (8, pp. 16^206} 11 j 23). An excellent discussion

of the principles and applications of ionization detectors

is offered by Dimick and Kigali (7) •

Before the development of gas chromatography, the

analysis of odors was a tedious, if not impossible, under*

taking. It usually involved fractionation by distillation,

followed by the examination of each fraction by physical

and chemical means* Since that time, both gas-solid

(adsorption) and gas-liquid {absorption) chromatographic

techniques have been employed in the analysis of a wide

variety of volatile compounds (1, 4, 15, 16, 17, 22)•

One of the most difficult problems to deal with in gas

chromatography has been that of chromatographing aqueous

solutions* Molecules of water and other strongly polar com-

pounds such as amines and alcohols are strongly adsorbed onto

common column packings and produce unsymmetric "tailing*1 of

the peaks, thus obscuring peaks of other compounds which may

be eluted shortly thereafter. As this investigation was

concerned with volatile compounds produced by an organism in

an aqueous medium, it was necessary 'to decide which of several

methods would be used to attempt the elimination of the inter-

ference due to the water#

Commonly used procedures for'water removal' by use of

drying agents, such as calcium chloride .and sodium sulfate,

and extraction of the unknots compounds, essentially free of

the water, wltfe organic solvents were considered briefly, but

were 'discarded. It was decided that these method® might

offer the possible disadvantages of either changing the com-

pounds chemically, or extracting or adsorbing some of the

components, but leaving others, thus yielding an inaccurate

©r an incomplete analysis.

Kung et al. (13) approached the sever®'water-tailing

problem by using an independently heated precclumn of calcium

carbide to convert the water vapor to acetylene.. The actylene

was then identified in all samples, including the controls,

as a single-, sharp- peak with no tailing-. This method was

also eliminated because of the high temperatures*-—-in excess

of 200 degrees Centigrade-—required in the heated precolumn

as such conditions would likely decompose or chemically change

the taste and odor compounds.

It has been suggested,' attd verified by lice (IS), that

the use of steam under pressure as the carrier vapor might

be a practical solution t© the water interference* Since

water would be constantly flowing through the eolumn and

detector, the water in the sample would be "zeroed out.*

10

Katurally thie necessitates the us# of apparatus' capable of

producing a continuous supply of steaa under constant

pressure# '

The relatively recent development of the hydrogen flame

ionization detector (?) has opened the way for what is

proving to be the beet method of controlling the water

problem in gas chromatography. The reason behind this success

is the fact that the flame detector does not respond to water,

arid, therefore, no water peak is present on the chromato-

graphic tracing.

Based on the reeowtendations and works of Baker (3), it

was decided to attempt to chromatograph directly the concen-

trated solutions and the vapors containing the taste and ©dor

compounds without prior treatment to remove the water* The

use of a hydrogen flame detector eliminated the need for

water removal# 1

The concentrated aqueous solutions and vapors were

analyzed directly on a Micro-fek GC-16G0 gas chromatograph

equipped with a hydrogen flame ionization detector* The

Instrument waa supplied with separate heating blocks, and

appropriate »onitoring systems, for the injection port, the

oven compartment containing the chrosat©graphic columns, and

the detector cell.

The recording system used waa a Brown recorder manu-

factured by Minneapolis Honeywell Corporation. This recorder

has appropriate mechanism© for producing a wide variety of

tracing speeds a® required.

XI

On tli© assumption • that the taste and odor compounds are

primarily polar compounds, especially aromatic amines and

aldehydes, it was decided to try chromatographic column

packings recommended for these groups. Literature on gas

chromatography (2, 5, 6, 9* 10, 12, 14, 19) showed that the

most widely used liquid eoatiags for the analysis of such

polar compounds, and for aqueous solutions, are classified as

polyethylene glycol® and commonly called Carbowax columns•

Other absorbent liquids which have been used with some

success are tricresyl phosphate, sorbitol sebaeate esters,

phthalic acid esters, fluorosilicone, silicone.oils, and

Armeen—-a mixture of high-boiling amines •

Two solid materials for gas chromatographic columns

which act as inert supports for the liquid .coatings haire

shown•promise in analysing solutions of polar compound®.

Teflon (14, 23) and alcoholic potassium hydroxide treated

firebrick (5, 21} haire been used to increase the retention

time of water and to reduce the tailing of peaks fro® water

and other polar compounds•

Two chromatographic columns were investigated in this

study. The first was a ten-foot long column of Carbowax 20-Jf

(20 per cent), using 60-00 mesh Chromosorb-P as the solid

support. Chromosorb-P is a common inert support manufactured

from diatomaceous earth. Carbowax 20-M is a polyglycol coat-

ing used for high-boiling amines and other polar compounds.

12'

' The second column studied was a Teflon coluan, nin# feet

in length, with no absorbent seating on the Teflon support»

The purpose ©£ this column was to investigate the possibility

of using Teflon to selectively adsorb, and thus separate, the

flavor and odor components«

Columns of fluorosilieone, tricresyl phosphate, and

another adsorbent column containing molecular sieve granules

were briefly investigated, but with incomplete and disap-

pointing result®* All coatings and supports investigated

were packed into one-fourth inch (outside diameter) copper

tubing with Swagelok fittings#

Temperature sequences in the injection port, column oven,

and detector cell were 'varied in the investigations of all ••••-•

columns in an effort to obtain, as near as possible, the :v

optimum conditions, the injection temperature was maintained,

after evaluation at several different settings, at ISO degrees

Centigrade in all tests to insure rapid vaporization ©f liquid

injections. Oven temperatures from 50 degrees to 200 degrees

Centigrade were examined. For good sensitivity without lose

©f baseline stability, the detector temperature was maintained

at 200 degrees Centigrade in all"studies* 1 , ' 1

Throughout this investigation, helium was used as the

carrier gas. Flow rates for the carrier gas were varied from

25 cc. per minute to 100 cc« per minute to determine that rat%

which would give the best separation and sharpness ©f peaks

eluted frop each column used. All flow rates were produced

with a tank pressure of 1*0 pounds per square inch.

13

The actinomycete cultures, a species of Streptomvces

labeled-in the laboratory as "Cooper*sB strain, were grown

in horizontal columns which ware previously sterilised with

ethylene oxide, The aediuia used, and sterilized by auto-

elaving, was as followsi ammonium nitrate |6 grams), sodium

chloride (3 grams), nutrient broth {12 grams), soluble starch

(2 per cent), Denton tap water {1500 milliliters)• Within

approximately seven t® tea days after inoculation with a

spore suspension, the odors given off by the culture were

swept from the column with a constantly flowing stream of

sterile air. The culture apparatus used was as described by

Silvey (20),

The ©dor and water laden vapors emitted from the column

were introduced into the gas chromatograph both directly as

gas samples and indirectly as concentrated liquid saaplee.

The gas samples were injected via a multi-port gas sampling

valve using samples of Z cc» and 5 cc. The liquid con-

centrates were obtained by freezing out with dry ice and

acetone the vapor coming from the culture columns. All

liquid samples, varying in volume from 2 to 5 microliters,

were introduced into the chronatograph by means of a 50-

microliter syringe.

At the time that this investigation was being terminated,

it was noted that several ponds in th© area surrounding

Denton, Texas, had developed strong ©dors similar to those

produced by the laboratory cultures of actinosycetes• The

14

principal odor was identified a® earthy. It • seemed desirable,:

therefore, to attempt a comparison of the chromatographic

tracings whieh could be obtained' from thee® surface waters t©

these of the odor concentrates from the culture columns.

T© produce concentrates fro® the natural surface water®,

another freeae-out technique was used# Approximately eight

liters ©f the water, collected frem a selected pond and

stored in a. polyethylene container for one to two days under

refrigeration, were filtered twice through glass filters.

The filtrate was placed in a cylindrical aluminum container

and surrounded by a jacket of ethylene glycol (antifreeze)•

The temperature of the surrounding antifreeze was lowered to

a minus one degree Centigrade, while the sample was stirred

by means of a glass rod extending approximately half the way

down along the central axis of the container and solution*

fhe constant stirring produced a cone-shaped aggitation area

in the aajsple. M the water slowly frost, fro® the periphery"

of the container inward, the other compound®, including the

©dorants, were concentrated eventually into the cone-shaped

portion at the canter. This concentrate, now about 500

milliliters, was removed with a pipette, placed in a clean

glass bottle and stored in a refrigerator. The earthy odor

was very pronounced in these concentrates. Samples of the

concentrates were chromatographed and compared with those

from the pure laboratory culture®.

15

Control columns or blanks, containing all ingredients

except the organism, were ehroaatographed to determine If the

results obtained might be due to compounds in the sterile air

lij|e» the medium, or the ethylene oxide used in sterilization,

the effluent vapors from th® control columns were chromato-

graphy directly and were also collected in liquid forra by

means of ac@tone~dry ice freese»out traps* This liquid was

then analysed in the gas chromatograph to compare with that

from the cultured experimental columns*

Finally, separate tracings were obtained directly from

the sterile air and from the ethylene oxide gas in an effort

to ascertain if extraneous peaks from the control column®

were dua to these elements.

REFERENCES CITSD II METHODS

1. Aerograph Research Motes, Biocheraicals with Electron MfinitZ, Wilkens Instrument and Research, Inc., California, 1964*

2. Altshuller, A. P. and C. A# Cle«sons, "Gas Chromatographic Analysis of Aromatic Hydrocarbons at Atmospheric Concentrations Using Flame Ionization Detection,* Analytical Chemistry, XXXIV {April, 1962), 466*472.

3. Baker, E. A., ttGas Chromatographic Analysis of Aqueous Solutions.* Materials Research and Standards. II (December, 1962); m ~ m .

4. Bauaann, F# and S. A. Olund, ^Analysis of Liquid Odorants by Gas Chromatography,n Journal of ChromatOEraphy. IX (March, 1962?, 431-435:

5* Bennett, C« E., 3. Dal Mogare, L. W. Safranski, and C, D. Lewis, "Trace Analyses by Gas Chromatography,® Analytical Chemistry, XXX (May, 1958), 898-902.

6. Bodnar, S» J. and S. J. Mayeux, "Estimation of trace and Minor Quantities of Lower Alcohols, Ethers, and Acetone in Aqueous Solutions by Oas Liquid partition Chromatography," Analytical Chemistry. XXX (August, 195&), 13^4^13^7•

7. Dimick, Keens F. and Louis A. Eig&li, Principles and Oas Chromatographic Applications of four itiaiatfon Detectors: Flpe, EIecteon gaj&urg, Cross Section, and Electron Mobility, Milkens Instrument and Research, inc•," Clifornia, paper presented at the Wilkens Gas

s&nsr®£Sfflinar'watckun8'Hew J,r"r' *. Heftmann, Erich, editor, Chromatography, New York,

Heinhold Publishing Corporation, '1961»

9. Hunter, I. R., V. H. Ortegren, and J. W. Pence, "Gas Chromatographic Separation of Volatile Organic Acids in Presence of Water*® Analytical Chemistry* XXXII (May, 1960), 6*2-6dV.

10* Johnson, D« E», %>« J• Scott, and A# Meister, "Gas*" Liquid Chromatography of Amino Acid Derivatives,M Analytical Chemistry, XXXIII (May, 1961), 669-673.

1 £

17

11. Keulemans, A. 1. M., Gas Chromatography, 2nd# ed., New fork. Relnhold PubliJlffii So,, 1959.

12* Knight, H, S., wGas-Liquld Chromatography of Hydroxyl and Amino Compounds.w'Analytical Chemistry. XIX (December, 195$/?

13. lung, J. T«, J. E. Whitney, and J. C. Cavagnol, "Analysis of Aqueous Solutions by Gas Chromatography," Analytical Chemistry* XXXIII (October, 1961), 1505-1507#

14* Landault, Catherine and Georges Guiochon, "Separation of Amines by Gas-Liquid Chromatography Using Teflon as a Support,M Journal of Chromatography% XIII (February^

15. MacKay, D. A. M., D. A, Lang, and M. Berdick, "Objective Measurement of Odor,w Analytical Chemistry, XXXIII (September, 1961),

16. Messner, A. B. M. Eosie, and P. A, Argabright, "Correlation of Thermal Conductivity Gell Response with Molecular Weight and Structure. Quantitative Qas

assess: 17. Ralls, J. W., "Rapid Method for Semiquantitative Deter-

mination of Volatile Aldehyde®, Ketones, and Acids. Flash Exchange Gas Chromatography," Analytical Chemistry. XXXII (March, I960), 332-336.

Id. Rice, J. K., Mgw Instrumental Methods of Wastewater Analyses, a report presented before the Fourth Industrial Water and Waste Conference, Texas Water Pollution Control Association*' Austin, Texas, 1964*

19. Bogozinski, 1., L. M. Shorr. and %. Warshawsky, "Gas Chropatographlc Analysis of Aqueous Alcohols," Journal of Chromatography. VIII (March, 1962), 429-432.

20. Sllvey, J. I* 0., James C. Russell, and David B. Redden, "Actinoiayeetes and Co«son tastes and Odors."Journal of American Water Works Association,.XLII Tlanuary ,' I W T T U O T : *

21. Smith, I. D. and E. D. Radford, "Modification of Qas Chromatographic Substrates for the Separation of Aliphatic Diamines." Analytical Chemistry. XXXIII (August, 1961), liso-msr

IB

22. Spancar, C. F.j F< Batiaann, and J. F. Johnson, ttQas Odorants Analysis by Gas Chromatography,M Analytical Chemistry, XXX (Saptaabar, 195$)» 1473-1474»

23. Szymanski. Herman A*, editor. Lectures on Gas Chromatog-ranhv 1962. He* lerk, ritmm Pr.ssT'lW.

CHAPTER III

RESULTS

Several physical parameters can be varied to affect the

efficiency of gas chromatographic columns. The most important

of these are the column length, the column temperature and

the carrier gas flow rate* An increase in column length

tends to prolong retention times of components and also,

\ therefore, produces better separation of peaks. Elution times

and separation efficiency vary indirectly with both column

temperature and carrier flow rate—-that is, increasing the

temperature or gas flow tends to decrease the retention times

of the components. It was felt that column® approximately

ten feet in length would be sufficient to evaluate the effi-

ciency and usefulness of a particular column without producing

retention times of unnecessary and Impractical durations. The

factors examined over relatively broad ranges to determine

approximate optimums were column or oven temperature and

carrier flow rate.

The Oarbowax 2Q-M column was evaluated over a temperature

range from 70 degrees to 200 degrees Centigrade. At each

temperature setting, the helium flow rate was varied between

50 cc. and 100 cc. per minute. At higher column temperatures

—-in excess ®f 100 degrees Centigrade——results were incon-

sistent, peaks were poorly separated, and significant tailing

20

©f many peaks >wa# produced* Sine® relatively large quantities)

five.microliter®, of aqueous concentrate were injected each

time, it is possible that, at the higher temperatures# small

amounts of the polyglycol coating were fluted off with each

injection, thus producing one or more of the large, tailed

peaks seen on injections of concentrates fro® odor-producing

and blank columns alike* This would naturally cause a de-

crease in column efficiency. Such a reaction was partially

verified by the decreased retention times of identical peaks

with each injection#

The optimum results were obtained with the Carbowax column

at $5 degrees Centigrade and the helium flow at 100 cc. per

ainute* Five microliter samples of each of three separate

fFeeze-out concentrates were chroiaatographed at these condi-

tions. A summary of the results, giving retention times of

the components, is shown in fable I, below.

TABUS I

RETENTION TIMES* OF THREE CONCENTRATES OK CARBOWAX 20-M AT 85® CENTIGRADE

HELIUM AT 100 cc. PER MINUTE

Concentrate from «larthyw Column

Concentrate from Uflarthy* Pond

Concentrate from Blank Column

2.9 • • • • # •

4*4 4.2 ...

• • • 6.1 ...

9.8 9.5 9.?

*in minutes.

21

Analysis of Table I reveals that each of the "earthy1*

concentrates produced two peaks not present in the tracing

of the blank column containing sterile medium. Each concen-

trate appeared to have one component common to both* This

component plus the first peak from the Cooper's concentrate

are believed to represent odor compounds. The second peak

from the pond concentrate may have been an odor component

also# It may have been, however, some extraneous compound

as it seems likely that several components other than odor

compounds would be present in such a sample•

It Is possible that the last peak, common to all tracings,

might be due to dissolved oxygen in the injected samples.

According to theory and literature (Szymanski, 1), the hydrogen

flame detector is not supposed to respond to oxygen# It Is

conceivable that a surge of oxygen through the detector cell

might cause the flame to burn more brightly, or at least

differently, thus initiating a signal of response# It is also

possible that the last peak is due to the elution of the water,

carrying with it a portion of the column coating.

Figure 1, Figure 2, and Figure 3 are tracings of the

Cooper's "earthy" concentrate, the "earthy" pond concentrate,

and the concentrate of a blank column of sterile media

respectively.

The nine-foot long Teflon column with no liquid coating

was tested first at 120 degrees Centigrade with the carrier

gas flowing at 100 cc. per minute. Under these conditions,

22

:: figs. 1, 2, 3. Oarbowax Column - 85°

:: inlet - 150° 0. :: Detector - 200® 0. o Helium - 100 oo,/ml&. "n 5 microliter samples

.QU. TT|"

"-"frtTf" ~r+

-i4-{ 4 -

-loj

4.4 279 4.4

Time {1 Inch = 1 minute)

Fig, 1—Concentrate from '•Earthy* Column ©n Carbowax

Time (1 inch = 1 minute)

Fig. 2—Concentrate from "Earthy" Fond on Carbowax

9.7 Time (1 Inch = 1 Alnute)

Fig, 3—-Concentrate fro® Blank Column on Carbowax

23

the three microliter odor concentrate Injections produced at

least two peaks, but the retention times were so short that

the separation was very poor# It was virtually impossible to

determine if several components had been condensed into the

two peaks noted. Based upon this data, it was decided that

lower temperatures and decreased gas flows should yield

better results.

At 70 degrees Centigrade and helium flow at 50 cc. per

minute, two peaks were again produced. Retention tines were

Omi and 1.2 minutes.

The optimum results appeared to be achieved with the

column temperature at 50 degrees Centigrade and the helium

flew at 25 oo. per minute, fable II summarises the results

obtained at these conditions.

TABLE XI

RETENTION TIMES* OF THRU CONCENTRATES TEFLON COLUMN AT 50° CEMTIGRADE HELIUM AT 25 CO. PER MINUTE

Concentrate from Concentrate from Concentrate from "Earthy" Column "Earthy0 Pond Blank Column

1*5 * • • 1.5 "

1.7 1.7 - * # #

2.1 2.1 - # ' # #

*in minutes

As indicated in Table II, the first peak, which is

present on tracings of concentrates from both experimental

24

and control culture columns but riot from the pond concentrates,

warfkown to be due t© dissolved oxygen, or possibly to

ethylene oxide. Gas injections of both oxygen and ethylene

oxide produced a strong, sharp peak with exactly the same

retention time as the first peak from the concentrates.

Since the culture columns were sterilised with ethylene oxide,

the presence of the gas in the concentrates could be accounted

for. The other peak, or possibly two peaks, from both ©dor

concentrates are believed to represent odorants.

Figure 4» Figure 5, and Figure 6 are representative

tracings of the Cooper*s "earthy" concentrate, the "earthy"

pond concentrate, and the concentrate from & blank column

respectively.

Investigations of the Fluorosilicone QF-1 {20 per cent)

on Chroaosorb-P and the molecular sieve columns produced un-

satisfactory results at all temperatures and flow rate# studied.

Temperatures between 75 degrees and 150 degrees Centigrade and

helium flow rates from 50 cc. to 100 cc. per minute were

examined. It was concluded that these columns are not sat-

isfactory for the solutions which were under investigation.

For all columns studied, gas samples of 2 cc. and 5 cc.

were also chromatographed. These samples consisted of the

odor-laden vapors which were omitted from the experimental

culture columns• Contrary to expectations, no significant

peaks were produced at any time by any of the columns when

gas samples were utilised.

25

figs. 4, 5, 6. teflon Column - 50° 0. 1;vt! Inlet - 150° 0. Detector - 200° 0. ^ Helium - 25 oo./mln.

; 3 minroliter samples >

1

2.1 1.7 1.5 Xlme (1 inch « 1 minute)

Fig. 4—Concentrate from "Earthy" Column on Teflon

YVHT

2.1 1.7 Time (1 inch * 1 minute)

Pig. 5-»Conc©ntrat© from wEarthylf Pond on teflon

lime (1 inoh = 1 minute)

Fig. 6—Concentrate from Blank Column on Teflon

REFERENCES CITBB IS RESULTS

1« Szymaiiskl. Herman Af. editor. Lectures on Gas Chromatog-raphy 1962. Mew York, Plenum Press, 1963•

26

CHAPTIB XT

DISCUS8I0H

Based upon the data presented in this report, it is

believed that gas chromatography can be an extremely useful

tool in the analysis of naturally occurring tastes and ©dors

in water supplies*

From the tracings obtained in this investigation, it is

believed that at least two odor components were present in

the "earthy" concentrates produced by s strain of actinomyeete

in the laboratory. Two peaks were obtained on both types of

gas chromatographic columns investigated, one column with

Carbowax 20-M (pe&yethylene glycol) as the liquid coating,

the other a Teflon column with no liquid partitioning agent.

The two peaks appeared to be absent in the liquid concentrate,

of blank culture columns containing sterile media. With these

odor components, it was found that, in general, better results

were obtained using lower column or oven temperatures and slow

to moderate carrier gas flows. These trends may be explained

on the basis that higher column temperatures might decompose

some of the component® yielding, therefore, inaccurate and

inconsistent results. The slower gas flows needed would seem

to be consistent with the low volatility of the odors. The

odors are extremely prominent and clear even when swelled at

temperatures near the freezing point of water.

zt

Additional procedures which should prove beneficial in

separation of ©dor components when used in conjunction with

gas chromatography are J (a) extraction of the components with

various organic solvents followed by chromatographic analysis

of the extracts, (b) operation of the gas chromatographic

column at temperatures approaching zero degree® Centigrade,

and (c) temperature-programmed gas chromatography.

Literature (3, 4, 7, 9, 12} indicates that temperature

programming should give significantly better separation than

the isothermal procedure used in this Investigation since,

in the former procedure, the oven temperature is raised

slowly and at a constant rate* Each compound is eluted as

its particular volatile temperature is reached.

After good separation of the odorants is achieved on a

particular column, identification of the constituent fractions

will likely be the next step. Methods which should be useful

in the classification and identification of the fractions

after their separation through the gas chromatograph are:

(a) collection of the separated fractions fro® the chromato-

graph and identification of each fraction by infrared

spectroscopy ©r by other methods ©f spectroscopic analysis

(6, &, 11, 13), and (b) utilisation of an electron capture

detection system as well as the hydrogen flame detector.

Some of the recent works (1, 2} conducted on gas chroma-

tography of tastes and odors has been carried out by means

of two-channel methods. By this procedure, two columns,

29

either of the same type or with different packings, are used

with two detectors, generally an electron capture end a

flame ionization. This allows the analysis'and detection of

a much wider variety of fractions and also gives a batter

indication &i to the chemical nature of compounds detected.

This same procedure has been used successfully in the anal-

ysis of insecticides and pesticides in natural water# ' (1, 5)#

By no means is it intended to presume that this inves-

tigation has exhausted the possibilities of gas chromatographic

columns or physical conditions which are applicable to the

separation of taste and odor compounds produced by micro-

organisms. This investigation, along with other work (3J0),

offers preliminary evidence to support the belief that gas

chromatography can be used in the separation, and eventually

the identification, of naturally occurring tastes and ©dors !

in surface waters. It is also indicated that adsorption, as

well as selective liquid absorption, gas chromatographic

column® can be used in the analysis of such flavor and ©dor

compounds.

REFERENCES CITED III DISCUSSION

1* Aerograph Research Motes, Biochemicals with glectron Affinity, Milkens Instrument and Research. Inc.. California, 1964.

2« Bonelli, 1. J., H. Hartmann, and K* P. Diaick, Qaa Chroma-tography Retention Times and Sensitivity Data for

im^lriwii)itl;i<ihwii itjiui iiui^iiinii^iiliiwiniiii wwriiiiiiiifiiiyjwi*iwi.iriMini, » i i y i w i | i M i i M p | i t » » i j w t „ mm m m m m i '*bw>iiiwwmmawiwiftiw.WIIIIMWWIIWMIIMIIW i j p w & E M t e .

Insecticides and Herbicides. Wilkens Instrument and Research, Inc., California, paper presented at th« Wilkens Gas Chromatography Seminar, Watchung, Sew Jersey, July 10, 1964*

Borfits, H., WA Simple Method of Temperature Programming for Gas Ghromatogrs

(October, 1< for Qaa Chromatography, "^Analytical Chemistry.

4* Dal Nogare, S. and C. E. Bennett, Programmed Temperature Gas Chromatography," Analytical Chemistry, XIX (June, 195$) 7 1157-11^.

5. Diaick, I. P. and J. W. Amy, Multichannel Systems for Gas Chromatography, Wilkeiis Instrument and Research. go., California, paper prac.nted at th« Wllkans Gas Chromatography Seminar, Watchung, New Jersey, July 10, 1964.

6. Gohlke, R. S., "Tirae-of-Flight Mass Spectrometry and Qas-Liquid Partition Chromatography,® Analytical Chemistry. XXXI (April, 1959), 535-541.

7. Keulemans, A. I. M., Gas Chromatography, 2nd ed.. New fork, Reinhold PubllihlBg Co.. ill?. P

B, Matthews, J. S., F. H. Burrow, and R. 1. Snyder, "Separa-tion and Identification of Cg Aldehydes. Use of Gas-Liquid Chromatography, Nuclear Magnetic Resonance, and Infrared Spectroscopy," Analytical Chemistry, XXXII (May, I960), 691-693.

9. Merritt, C. and J. T. Walsh, "Programmed Cryogenic Tem-perature Gas Chromatography Applied to the Separation of Complex Mixtures," Analytical Chemistry, XXXV (January, 1963), llO-llj;;

10. Eice, J. K«, New Instrumental Methods of Waste-Water Analyses, a report presented before the Fourth Industrial Water and Waste Conference, Texas Water Pollution Control Association, Austin, Texas, 1964.

. . 31

11. Snouse, T« H. and S. S. Chang, "Micro-Fraction Collector for Gas Chromatography,« Journal of Chromatography, XIII {January,, 1964), 244-246..,

12. Sullivan, J# K« and J. f. Walsh, "Improved Separation in tea Chromatography by Temperature Programming,» Analytical Chemistry. XXXI (Itotrember, 1959)> 1826-182$.

13. Thomas, P. *J» and J. L. Dvyer, "Collection of Gas-Chromatographic Sffluent8 for Infrared Spectral Analysis,1* Journal ef Chromatography, XXII (February, W&k), 366-371 •

SUMMARY AMD CONCLUSIONS

1. The use of gas chromatography in the separation of

naturally occurring tastes and odors was investigated. The

concentrated aqueous solutions of the ©dor compounds were

analyzed without previous treatment for water removal,

2. two gas chromatographic columns were investigated

and found t© be satisfactory in the separation of the odor

component®• One colunn contained Garbowax 20-M (polyethylene

glycol) as the liquid partitioning agent. The second column

contained Teflon as an adsorbent support with no liquid

coating. A hydrogen flame ionization system was used as the

detector.

3. fhe concentrated odor solutions were produced by

freeze-out techniques. One concentrate was from the odor-

laden vapors fro» a strain of an actinoraycete cultured in the

laboratory, fhe other solution was a concentrate of water,

collected from a stock pond. Both concentrates were charac-

terised by strong "earthy" odor®.

4. Two peaks, either or both of which nay prove t© be

odorants, were noted on tracings of both concentrates from

both columns examined#

5. Based upon data presented in this report, it is be-

lieved that gas chromatography can be a very useful tool in

the analysis of naturally occurring tastes and odors in water

supplies.

BIBLIOGRAPHY

Beoks

Heftmann, Krich, editor, Chromatography, lew Tork, Reinhold Publishing Corporation, 1961•

Xeulemanf, A. I. M., Gas Chromatography, 2nd. ed«, Mew Tork, Reinhold Publishing Co., 19$9.

Szymanski, Herman A., editor. Lectures on Gas Chromatography 1962, Hew Tork, Plenum Press, 1963.

Articles

Altshuller, A. P. and C. A. Clemmons, "Gas Chromatographic Analysis of Aromatic Hydrocarbons at Atmospheric Concentration® Using Flame Ionization Detection,* Analytical Chemistry, XXXIY {April, 1962)» 466-472.

Baker, 1. A#, "Gas Ohrostat©graphic Analysis of Aqueous Solutions,Materials Iggeareh and Standards, II (December, 1962), J"L

Bauaann, F. and S. A. Olund, "Analysis of Liquid Odorants by Gas Chromatography,* Journal of Chromatography, II (March, 1962), 431-43C

Bennett, C. 1., S. Dal Nogare, L. W. Safranski, and C. D. Lewis, "Trace Analyses by Gas Chromatography,n Analytical Chemistry. XXI (May, 195$), *98-902.

Bodnar, S. and S. J. Mayeux, "Estimation of Trace and Minor Quantities of Lower Alcohols, Ethers, and Acetone in Aqueous Solutions by Gas Liquid Partition Chromatography,» Analytical Chemistry. XXX (August, 1958)» 1384-1387.

Borfitz, I., *A Simple Method of Temperature Programming for Gas Chromatography,* Analytical Chemistry> XXXIII (October, 1961), 1632.

Dal Nogare, S. and C. E. Bennett, "Programmed Temperature Gas Chromatography,B Analytical Chemistry. XXX (June, 195©), 1157-1158.

O *%

34

Gaines, H. D. and E» P. Colline, "Volatile Substances Produced by Streptomyces odorifer," Lloydia. XXVI (December, 1963), 247-253•

Gohlke, 1. S.» MTime~©£-Fli ght Mass Spectrometry and Gas-Liquid Partition Chromatography." Analytical Chemistry. XXXI (April, 1959), 535-541.

Hunter, I# 1., V. I# Ortegren, and J. V, Pane®, "Gas Chro-matographic Separation of Volatile Organic Acids in Presence of Water** Analytical Chemistry. XXXII , m&n 1960), 682-w:

James, A. T. ahd A. J. Martin, f,das~Liquid Partition Chro-mt^graphy,* Biochemical Journal* L (June, 1952),

Johnson, D. £., S. J. Scott, and A. Meister, ttGas~Liquid Chromatography of Amino Acid Derivative*.• Analytical Chemistry. XXXIII {May, 1961),' 669-673.

Knight, H. S., "Gas-Liquid Chromatography of Hydroxyl and i%i n» w• | fiiias**jy&qiiici viirciaEt griipiiy ©i

*Dec«»lb«?P01956i" jSSfo^fir chwl"trT-

Rung, J* T., J. K« Whitney, and J. C. Cavagnol, Analysis of Aqueous Solution® by Gas Chromatography," Analytical Chemistry. XXXIII {October, 1961), 1505-1507.

Landault, Catherine and Georges Quiochon, ^Separation of Amines by Gas-Liquid Chromatography, Using Teflon as a Support,* Journal of Chromatography, XIII (February, 1964), 327-336,

Laughlin, Harold F., ^Palatable Level With the fhreshhcid Odor Teat,* Taste and Odor Control Journal. XX (August, 1954)# 1-8.

Lendall, Harmon I., "A Comprehensive Survey of the Taste and Odor Problem,* Taste and Odor Control Journal. XII (June, 19461, 1-17.

MacKay, D. A. M«, D. A. Lang, and M. Berdick, *Objective Measurement of Odor,** Analytical Chemistry. XXXIII (September, 1961), 1369-1174. —

Matthews, J. S., F. H. Burrow, and R. I. Snyder, Separation and Identification of Cg Aldehydes. Use of Qas-Liquid Chromatography, Nuclear Magnetic Resonance, and Infrared Spectroscopy,* Analytical Chemistry. XXXII (May, I960), 691-693.

35

Merritt, C. and, J. t. Walsh, Cryogenic J « P ^ u r .

110-113.

Meesner, A# 1., D. M.Eosie, and P.. A• *!!?Grre~ lation of fhermal Conductivity Cell Eesponse with Molecular Weight and Structure. Chromatographic Analysis," Analytical Chemistry, XXXI (February, 1959)» 230-233*

lulls, J* ¥•» "Rapid Method for Semiquantitative^Determina-tion of Volatile Aldehydes, Ketones, and Acids* Flash ' Exchange Gas Chromatography,n Analytical Chemistry, XXXII {March, I960), 332-33&.

Rogozinski, M., K M. Shorr, and Z. Warshawsky, "Gas Chro-matographic Analysis of Aqueous Alcohols," Journal of Chromatography* ¥I2X (March, 1962), 429-432.

Rohlich, George A. and William B. Sarles, «Ch®mical Compo-sition of Algae and Its Relationship to fast® and Odor," fast® and Odor Control Journal, XVIIX (November, 1949)* 1=7:

"The Chemistry of Organic Compounds Responsi ble for Pastes and Odors," faste and Odor Control Journal* XV (June, 1949)> 1-5«

Silvey, J. I* 0., James C. Eussell, and 5 a S d R#«R?ddS!S "Aetinemycetes and Common Tastes and Odors," ,j?urygL$£

Icim Water Works Association, XUZ Uwswy>

Silvey, J. K. 0., James C. Russell, D. E. R e d ^ » , W. McCormick, "Actinoiaycetes and Common Testes and

TL°^reE2^^)^4^=0ir i MaS££

Smith, 1. D. and E. C. Eadford, "Modification of Qas Chro-matographic Substrates for the Separation of Aliphatic Diamines," Analytical Chemistry« XXXIXI (August, 1961), 1160-1162.

Smouse, 7. H. and S. S. Chang, "Micro-Fraction Collector for ©as Chromatography,* Journal of Chromatography, XXIZ (January, 1964)» 244-246.

Spencer, C. P., F. Baumann, and J. P. Joh^on, "Gas Odorsuats Analysis by Qas Chromatography,w Analytical ChfWBjstr , XXX (September, 195**) > 1473-1474.

36

Sullivan* J. H. and J. T. Walah, "Improved Separation in Gas Ohronatography by TenperatuiPB Programlog," Analytical Chemistry* XXII (November, 1959)» 1826-1828.

•Thomas, f* and J. X. "Collection of Gas-Chromatographic Effluants for Infrared Sp«etral AnalMie," Journal of Chromatography» XXIX (Februaryt 196V)* 366-371•

.,. Reports

Aerograph Research Hotee, Biochemicals with VlXkms lEstnja.#nt- andEtseareii, Xac *, California, lyok?

Bonelli} S• J«, H. Hartm&nn, and K» P * , S&M, SHSffi.*"

&ssEfe&isi>R£!Sh!£ S ^ § i i « £ S ^ s « r r . r h ^ : % ~ J . r . . y , July 10f 1964*

Blnleki K. P. and J. W. toy, Multichannel gZ&giB,|g£ gSS n -rnmat,Qg;raphY. Wilkens Instrument • ana Research, Ine *,

. California, pap^r presented at.the Wilkens_Gas Chroma-tography Seminar» Wat chung, New Jersey, July 10 , 1964*

is detectors t

>PlJLltY« JIIXK9UB, i i w v r u w f l b ***" * . . T Z t w * t* 1. California, paper presented at the Milkens GasChroma-tography Seminar, Watchung, Mew Jerrgsy, July 10, 1964<

Rice, J» K., and a report presented before the rourta inaus rxax wau«i a Waste Conference, Texas Water Pollution Control Associ-ation, Austin, Teams, 1964*

U&fmblished Materials

Bill. W. S.. '•The Chemical Compounds Produced by Actinoiiycetee and Their Relation to Tastes and Odors in a Water Supply," unpublished master's thesis* Department of Biology, North Texas State University, Benton, Texas, 1951.

McCormick, William C., "The Cultural, Physiological, Morpho-logical and Chemical Characteristics of an Antinomycete from lake Waco, Texas," unpublished master's thesis, Department of Biology, Morth Texas State University, Benton, Texas, 1954*

37

Pipes, Wesley 0., Jr., nAn Investigation of Matufrally Occurring fastis and. Odors fro® Fresh Waters," unpublished master's thesisv Department of Biology, North Texas State Uni-versity, Denton, Texas, 1951.

Pipes, Wesley 0#, Jr., *An Investigation of Naturally Occurring tastes and Odors from Fresh Waters*",unpublished master's thesis, Department of Bifclogy, North Texas State Univer-sity, Denton, Texas, 1951, citing Janes Churchill, Report on $he Cholera Outbreak in the Parrish of St. aaies,' fistiinlster, dping thfTutumn on lgjj, presented to the Vestry by the Cholera Inquiring Cowitttee, July, 1855.