Reactions of chlorine disinfectant with organics adsorbed on granular activated carbon

MISSING IMAGE

Material Information

Title:
Reactions of chlorine disinfectant with organics adsorbed on granular activated carbon
Series Title:
Florida Water Resources Research Center Publication Number 70
Physical Description:
Book
Creator:
McCreary, John J.
Batsel, Kurt R.
Rivera, Javier R.
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Notes

Abstract:
The reactions of aqueous free chlorine with two organic molecules were examined before and after adsorption on granular activated carbon. The molecules were both typical of subunits of naturally-occuring humic and fulvic acids. Reactions were conducted at chlorine/carbon ratios and at carbon column flowrates typical of those encountered in water treatment practice. It was observed that carbon catalyzed the hydroxylation of the aromatic ring in all reactions, even when chlorine was not present in the influent to the column. It was also observed that large molecular weight products could form, presumably via a free-radical mechanism at the carbon surface. The products of all reactions were determined by capillary GC/MS and in some cases, high resolution GC/MS was employed for additional structural information. Reaction of one organic at a higher pH produced similar products to those found at neutral pH. This work was then extended to the chlorination of a fulvic acid water sample on activated carbon. The carbon at the influent end of the column contained compounds that were mutagenic in the Ames test, and had innumerable phthalates and fatty acid methyl esters.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution and holding location.
System ID:
AA00001546:00001


This item is only available as the following downloads:


Full Text














Publication No. 70

THE REACTIONS OF CHLORINE DISINFECTANT WITH ORGANIC
ADSORBED ON GRANULAR ACTIVATED CARBON

By

John J. McCreary
Kurt R. Batsel
Javier R. Rivera


Department of


Environmental Engineering Sciences
University of Florida
Gainesville


l 4














TABLE OF CONTENTS


Abstract iv
Acknowledgements v


Chapter Page


1. Literature Review 1


2. Materials and Methods 3

2.1. Activated Carbon 3
2.2. Water Purification 3
2.3. Chlorine Solution 5
2.4. Organic Compounds 5
2.5. Column and Reactor Configurations 8
2.6. Extraction of the Granular Carbon Bed 8
Samples
2.7. Gas Chromatographic Mass Spectral 8
Analysis


3. Results and Discussion

3.1. Aqueous Chlorination of Vanillic Acid 14
3.2: Chlorine/Adsorbed Vanillic Acid 16
Reaction
3.3. Aqueous Chlorination of Ferulic Acid 28
3.4. Ferulic Acid/Chlorine/Carbon Reaction 40
3.5. Aqueous Chlorination of Ferulic Acid 54
at pH = 10
3.6. Ferulic Acid/Chlorine/Carbon Reaction 58
at pH 10
3.7. Aqueous Chlorination of a Natural 58
Water Sample















Chapter


4.


References


Appendix


Conclusions


Page


69


71


74














ABSTRACT


The reactions of aqueous free chlorine with two organic molecules
were examined before and after adsorption on granular activated carbon.
The molecules were both typical of subunits of naturally-occuring humic
and fulvic acids. Reactions were conducted at chlorine/carbon ratios
and at carbon column flowrates typical of those encountered in water
treatment practice. It was observed that carbon catalyzed the hydroxyl-
ation of the aromatic ring in all reactions, even when chlorine was not
present in the influent to the column. It was also observed that large
molecular weight products could form, presumably via a free-radical
mechanism at the carbon surface. The products of all reactions were
determined by-capillary GC/MS and in some cases, high resolution GC/MS
was employed for additional structural information. Reaction of one
organic at a higher pH produced similar products to those found at neutral
pH. This work was then extended to the chlorination of a fulvic acid
water sample on activated carbon. The carbon at the influent end of the
column contained compounds that were mutagenic in the Ames test, and had
innumerable phthalates and fatty acid methyl esters.














ACKNOWLEDGEMENTS


The authors are indebted to Carl. J. Miles, the operator of the
gas chromatograph/mass spectrometer, for his advice on processing
the samples. We also wish to acknowledge the help of the glassblowing
shop in the Chemistry department for the preparation of the columns
and Dr. Roy King for the high resolution GC/MS results.














CHAPTER 1. LITERATURE REVIEW


It must be assured that activated-carbon adsorption is a safe
and effective method for removing organic compounds if it is to be
widely implemented to treat drinking water. It should be established
that, when it is used, no potentially toxic organic will be formed on
the carbon surface and passed into the effluent from the bed. In most
water treatment systems, chlorine added to disinfect the water supply
will come in contact with activated carbon and may react with the acti-
vated carbon or organic adsorbed on the carbon. The end products of
these reactions need to be carefully examined to determine whether
potentially harmful compounds are formed and released to the water
being processed (McCreary et al., 1982).

Several researchers have demonstrated that activated carbon can
promote reactions in aqueous solution. Reactions cited have included
the catalytic oxidation of inorganic substances such as sodium nitrite,
potassium arsenite, sodium sulfite, and potassium ferrocyanide (Puri
et al., 1958a); stannous chloride (Larson and Walton, 1940; Puri et al.,
1958b; King, 1936; Farmer and Firth, 1924); and potassium urate (Bente
and Walton, 1943).

Several workers have cited the catalytic oxidation of organic
compounds in aqueous solution by activated carbon. These have included
the oxidation of oxalic acid from aqueous solution (Rideal and Wright,
1926); and the oxidation of n-butyl mercaptan to butyl disulfide (Ishizaki
and Cookson, 1974). McCreary et al. (1981) have demonstrated that the
reaction of chlorine with adsorbed humic substances did not produce large
quantities of halogenated compounds. However, extraction and analysis
with GC/MS of the influent to the column and the activated carbon showed
several dihydroxy and chlorinated dihydroxybenzenes on the carbon that
were not present in the influent. Similar work with vanillic acid
(McCreary et al., 1982) and 2,6 dimethyl phenol (Voudrias et al., 1982)
have demonstrated that hydroxylation of the aromatic ring appears to be
a general reaction associated with carbon-adsorbed organic-chlorine
interactions.

The catalytic oxidation of various substances on carbon has fre-
quently been related to surface functional groups of the carbon containing
oxygen. The oxidation of inorganic salts was found to correlate with the
acidic surface oxides on the carbon (Puri et al., 1957, 1958a; King,
1936; Larson and Walton, 1940); however, tie catalytic decomposition of
hydrogen peroxide proceeded most rapidly on carbon that had an alkaline









surface (Puri et al., 1958b; King, 1936; Larson and Walton, 1940; Bente
and Walton, 1943).

The oxidation of organic materials on activated carbon has been
explained using a mechanism involving chromene-type oxygen groups (Garten
and Weiss, 1957) and by quinones (Ishizaki and Cookson, 1974). In each
case, there was little experimental evidence for these interpretations.
Chen et al. (1983) evaluated the role of various surface functional groups
in the reaction of indan with chlorine dioxide on carbon. It was determined
that free radicals on the carbon surface were responsible for the production
of chloroindan. When the free radicals were inhibited by reaction with
diazomethane, the reaction was reduced. A similar reduction did not occur
when carboxylic acids or phenolic functional groups were reacted with
appropriate blocking agents.

Several investigators have demonstrated that metal ions in activated
carbon can significantly enhance catalytic reactions. Rideal and Wright
(1926) demonstrated that an Fe-C complex was 50 times more active than
the carbon surface alone, and an Fe-C-N complex was 800 times more active
for the aqueous oxidation of oxalic acid. Larson and Walton (1940) noted
that the decomposition of H202 was enhanced by copper sulfate and especially
ferric sulfate. Bente and Watton (1943) found that the rate of catalytic
decomposition of H202 for a variety of nitrogenous carbons was reduced
with decreasing ash content, although specific metals were not determined.
Other work has shown that Mg+2 on carbon could support the oxidation of
hydrogen sulfide (Swinarski et al., 1978), and that iron and copper promoted
the oxidation of butyl mercaptan (Ishizaki and Cookson, 1974). Recently,
however, Chen et al. (1983) showed that a specially prepared sugar-based
carbon had catalytic properties although it had an extremely low metal
content.

A National Academy of Sciences report (1979) indicated that the
influence of metal ions on the promotion of catalytic reactions is of
particular interest since the metal content should increase after carbon
is regenerated. Thus, carbon that has passed through several regeneration
cycles may have an enhanced reactivity.

The purpose of this study was to extend the organic-chlorine reaction
on activated carbon to additional compounds. In addition, the influence
of pH was investigated. Finally, preliminary work was done to relate the
products formed in the chlorine-model compound work to those products
formed in the reaction of natural humic acids with chlorine on carbon.














CHAPTER 2, MATERIALS AND METHODS


2.1. Activated Carbon

The carbon used in all column experiments was a bituminous base
granular activated carbon (GAC) (F-400, Calgon Corp., Pittsburgh, PA).
It was prepared by sieving to the desired size fraction (50 x 100 U.S.
Standard Mesh), washed with deionized water, dried, and baked at 1050 C
for 3 weeks to remove impurities.

The ash and metal content were determined for F-400 and for another
granular activated carbon (HD-3000, ICI America, Inc., Wilmington, DE)
prepared in the same manner. The results are shown in Table 1.

Hydrodarco-3000 showed a higher percent of ash compared to F-400.
Other manufacturers of activated carbons report percentages lower than
8% (Culp et al., 1978). Filtrasorb-400 had a smaller ash content than
HD-3000 and also had smaller concentrations of iron and copper. The metal
content in activated carbon has been related to the catalytic activity of
the carbon in several reactions. Rideal and Wright (1926) demonstrated
that an iron-carbon-nitrogen complex was 800 times more active than the
original carbon for the oxidation of oxalic acid in aqueous solution.
Larsen and Walton (1940) concluded that the decomposition of H202 by a
carbon was enhanced by the presence of copper sulfate, and especially by
ferric sulfate. However, recent work by Chen et al. (1983) suggested
that metals were not necessary for the incorporation of C1 from C102 into
an organic compound (indan) at the carbon surface.

The F-400 and HD-3000 carbons were Soxhlet extracted to determine
organic contaminants on the carbon surfaces. Figures 1 and 2 show total
ion current chromatograms of the extracts of F-400 and HD-3000, respec-
tively. Filtrasorb-400 appeared to have one large contaminant that was
determined to be the ester of a fatty acid (adipate). The Hydrodarco-3000
was much more contaminated than the F-400 carbon, possibly as a result of
adsorption of organic from soil simultaneously dried in the same oven.
Filtrasorb was chosen to continue the column experiments due to the inability
to remove the contaminants from the HD-3000.

2.2. Water Purification

The water used to prepare the solutions in the initial experiments
was purified in order to remove most of the organic contaminants. Distilled


























Table 1. Ash and Metal Analysis for F-400 and HD-3000


Type of GAC F-400 HD-3000

Ash % 6.20 14.0

Copper % 0.002 0.003

Iron % 0 .14 1.08








water was pumped through a Milli-Q Water Purification System (Millipore
Corporation, Bedford, MA) which consisted of one ionic exchange resin
followed by two consecutive activated carbon cartridges. This water was
collected and again pumped through a 10 cm column of F-400 GAC. The
effluent water was collected and stored in 45 liter carboys.

Fifty liters of the effluent water were passed through a 3 cm F-400
GAC column and an extract of the carbon was analyzed by GC/MS. A total
ion current chromatogram is shown in Figure 3. This chromatogram demon-
strates that the water was sufficiently clean to prepare the solutions.
The impurities in the water-carbon extract appear to be identical to those
on the F-400 carbon (Figure 4).

Problems were encountered with distilled water later in the experiments
due to the failure of the laboratory still. This was remedied by using tap
water preextracted through an F-400 carbon filter as described above.


2.3. Chlorine Solution

High purity chlorine gas (Matheson, Searle Medical Products USA, Inc.,
New York, NY) was bubbled into distilled-deionized water containing several
grams of NaOH per liter. The concentration of this solution was checked
using the DPD Method (Standard Methods, 1980). An appropriate amount of
stock solution was added to purified water containing a 0.001 M phosphate
buffer to achieve a chlorine solution of required concentration for the
column runs. The chlorine influent bottle was covered by dark plastic and
capped with aluminum foil.


2.4. Organic Compounds

Two phenolic acids were used in the experiments: vanillic acid
(3-methoxy, 4-hydroxy benzoic acid) (Sigma Chemical Co., St. Louis, MO),
and ferulic acid (3-methoxy, 4-hydroxy cinnamic acid) (Aldrich Chemical
Company, Inc., Milwaukee, WI). Vanillic acid is known to be a degradation
product of humic materials present in natural waters (Christman et al.,
1978); it has been previously studied in experiments of this type (McCreary
et al., 1982) and has been shown to yield products that can be characterized
by gas-chromatographic mass spectrometry (Larson and Rockwell, 1977, 1979).
The second compound, ferulic acid, was chosen due to its structural simi-
larity to vanillic acid, the fact that it contained an unsaturated alkyl
chain, and that its products in aqueous chlorination reactions have been
determined (Norwood et al., 1980).

Vanillic acid was purified by recrystallization in a mixture of
water and methanol. Ferulic acid was initially used without further puri-
fication and was subsequently purified by recrystallization in a water/
methanol mixture. Reaction solutions were made by dissolving carefully
weighed amounts of each compound in 100 ml of water made alkaline with


























Figure 1. Total Ion Current Chromatogram of an Extract of F-400 (GAC).


Figure 2. Total Ion Current Chromatogram of an Extract of HD-3000 (GAC).





















.-_., __ / I.__.*' ,. ,_ ._ ,__


Figure 3. Total Ion Current Chromatogram of an Extract of the Purified Water,







F-400 (GAC)










Purified Water Extract

_, A ____ / _

Figure 4. Comparison of the Total Ion Current Chromatogram of an Extract
of F-400 (GAC) with the Purified Water Extract.









several pellets of NaOH. This solution was added to a carboy containing
45 liters of purified water (distilled or tap) buffered with 0.001 M
phosphate solution. The carboy was capped with aluminum foil.


2.5. Column and reactor configurations

Several different columns were used in the adsorption experiments.
A typical assembly is shown in Figure 5. The entire apparatus was constructed
of glass, teflon and stainless steel to minimize contamination of the
solutions. The phenolic acid and chlorine were each pumped separately through
metering pumps (Model G-50, Fluid Metering, Inc., Oyster Bay, NY) and entered
the bottom of the columns through a stainless steel T-union (Swagelok,
Crawford Fitting Co., Cleveland, OH). Flowrates were adjusted so that flow
through the columns was maintained at 2 gpm/ft2 with a 2 minute contact time
before the solution came in contact with the carbon bed. In several cases,
smaller columns were used to allow slower flow rates, so that extended
column runs could be performed with the same amount of solution. Columns
were prepared by placing approximately 10 cm of carbon between pyrex wool
plugs. A blank column was prepared for each experiment with only pyrex wool
in the column and the flow rate adjusted such that a 2 minute reaction time
was maintained. A typical blank column is shown in Figure 6. Compounds
present in an extract of carbon after adsorption but not present in the blank
solution represented compounds formed only in the presence of activated
carbon. Initially the effluent of the blank reactor was collected, reduced
with an excess of sodium sulfite, adjusted to the pH of the influent solution
and pumped onto an F-400 granular activated carbon bed. After the solution
was passed, the bed was dismantled and Soxhlet extracted as described below.
The experimental arrangement is shown in the schematic in Figure 7. Later,
it was observed that the carbon used for the blank solution was producing
catalytic products. The experimental procedure was then altered to collect
blank solution in a 4 liter brown glass bottle, reduced with Na2SO3, adjusted
to approximately pH 2 and concentrated by MeC12 liquid-liquid extraction.


2.6. Extraction of the Granular Carbon Bed Samples

Approximately one gram samples of carbon were placed in a Soxhlet
extraction apparatus. One milliliter of methanol (Burdick and Jackson,
distilled-in-glass, Muskegon, MI) was added to the wet carbon and the sample
was Soxhlet extracted with methylene chloride for 12 hours. The extract
was dried with anhydrous sodium sulfate (Na2SO4) and concentrated to
approximately 1 ml in a Buchi Rotary Evaporator (Model R110, Brinkmann
Instruments, Inc., Westbury, NY).


2.7. Gas Chromatographic Mass Spectral Analysis

The concentrated extracts were methylated with diazomethane generated
from N-methyl-N-nitroso-p-toluenesulfonamide (Diazald) using the Aldrich





1/8" Teflon Tubing


15 cm
Length

1.9 cm
Diameter


Chlorine
Solution


Swagelok 1/8" Stainless
Steel Union


-j- Teflon Stopcock

--- Pyrex Glass Wool


--- -50 x 100 Mesh F-400 Carbon














-- Pyrex Glass Wool


24/40 Ground Glass Seal




- -- Teflon Stopcock



Swagelok 1/8" Stainless
7/ Steel T-Union


,. Reactant Organics


Figure 5. Carbon Column Configuration (after McCreary et al., 1982).





















1/8" Teflon Tubing


Chlorine
Solution


Swagelok 1/8" Stainless
Steel Union


-- Teflon Stopcock

. Pyrex Glass Wool






24/40 Ground Glass Seal




- Teflon Stopcock



Swagelok 1/8" Stainless
SSteel T-Union


.-- Reactant Organics


Figure 6. Blank Reactor Column Configuration (after McCreary et al., 1982).








Experimental Arrangement of the Carbon-Chlorine Reactions


Procedure #1


HOC1


.organic


HOC1


2 min reaction time


carbon



Soxhlet extraction

CH2N2


GC/MS


reduced with
Na2S03

1
carbon

Soxhlet extraction

CH2N2

GC/MS
GC/MS


Procedure #2


.organic


2 min reaction time


Soxhlet extraction
1
CH2N2

GC/MS


reduced with
Na2SO3

pH adjusted to 2

extracted with
CH 2C12
CH2N2

GC/MS


Figure 7.









Diazomethane Generating Kit (Aldrich Chemical Co., Milwaukee, WI). Ten
microliters of D10 anthracene (5.1 mg/ml) or D8 naphthalene were added
as an internal standard. The methylated samples were analyzed by injecting
2 microliters of each sample into a 30 meter DB-5 Fused Silica Capillary
Column (J & W Scientific, Rancho Verde, CA) interfaced to a 5985 B GC/MS
(Hewlett Packard, Palo Alto, CA). Typical gas chromatograph and mass
spectrometer conditions are shown in Tables 2 and 3 respectively. The
Eight Peak Index of Mass Spectra (1974) and the EPA/NIH Mass Spectral
Data Base (1978) were used as an aid in the identification of the peaks.





















/














Table 2. Chromatographic Conditions for the Analysis of the Samples


1. Instrument: 5985 B GC/MS (Hewlett Packard, Palo Alto, CA)

2. Column: 30 meter DB-5 Fused Silica Capillary Column
(J & W Scientific, Rancho Verde, CA)

3. Interface: Direct interface to mass spectrometer

4. Carrier Gas: Helium (Zero Grade)

5. Flow Rates: Column Flow: 2cm/min

6. Detector: Electron Multiplier

7. Temperature Program:

Temp. 1...3000
Time 1....5 minutes
Rate .....4.0C/min
Temp. 2...2800C
Time 2....30 minutes











Table 3. Mass Spectrometer Conditions for the Analysis of the Samples


Repeller (volts) 6.29

Drawout (volts) 36.

Ion focus (volts) 189.

Entrance Lens (MV/AMU) 105.

X-ray (volts) 96.

Emission (UA) 300.

Electron Energy (EV) 70.

Electron Multiplier Voltage 1800.

AMU Gain 160.

AMU Offset 67.

Mass Axis Gain 1.01037

Mass Axis Offset -0.424866

Ions: Positive

Actual Source Temperature: 2020C


Source: Autotune Final Report, 6/2/82

13














CHAPTER 3. RESULTS AND DISCUSSION


3.1 Aqueous Chlorination of Vanillic Acid

The aqueous chlorinati-on of vanillic acid was studied by reacting
the compound with chlorine for 2 minutes, followed by reduction with sulfite
and collection of the reaction products on carbon. The experimental para-
meters for the aqueous reaction are presented in Table 4. A total ion
current chromatogram of the F-400 methylated extract is presented in
Figure 8 and a list of tentatively identified compounds from the aqueous
chlorination of vanillic acid is shown in Table 5. Due to the complexity
of the products, compounds are identified by functional groups and, to
avoid ambiguity, the compounds are named and presented as the methylated
derivatives found in the extract.

Two chlorine substituted phenols were identified in the extract.
These were chloro methoxy phenol (1) and dichloro methoxy phenol (3). The
fully methylated derivative of (3), dichloro dimethoxy benzene (2) was also
present in the extract.

A series of methylated carboxylic acids were the major products of
the reaction. Methyl, dimethoxy benzoate (5) and methyl, methoxy hydroxy
benzoate (4) were detected in appreciable amounts. Two chlorine substituted
benzoates were also found: methyl, monochloro dimethoxy benzoate (6) and
methyl, dichloro dimethoxy benzoate (9). Compounds 10, 11 and 12 were of
high molecular weight (352,33a8,386). Possible structures are discussed
in Appendix A.

The % areas of the compounds are listed in Table 5 for comparison
purposes. These abundances can only be roughly related to the concentration
of the compounds in the reaction since the area is related to the number
and intensities of fragment ions in the mass spectrum. More than 60% of
the area was attributed to unreacted methylated vanillic acid. The single
chlorine substituted product, methyl, monochloro dimethoxy benzoate (6)
represented more than 25% of the total compound area in the extract and the
dichloro substituted product, methyl, dichloro dimethoxy benzoate (9)
represented only a minor percent of all the products formed. The monochloro
and dichloro methoxy phenols were present in an abundance of less than 3%
each. The low abundance of chlorinated phenols could be due to the ineffi-
ciency of extraction of phenols from the activated carbon as reported by
Chriswell et al. (1977).














Table 4. Experimental Parameters for the Vanillic Acid/
Chlorine Blank Reactor Column


Column Diameter, cm

Column Length, cm
Influent Vanillic Acid Conc., mg/l

Influent Chlorine Conc., mg/l
Influent pH
Flowrate, 1,/m2-min
Reaction time, min

Reaction volume, ml

Total volume passed through bed, 1


4















2 3
1


5








IL


1.9
15

2.6

10.0
6.0
68.8

2.0
38.6
18.5


8
7



S 9 10
*'J L' 'h-


Figure 8. Total Ion Current Chromatogram of an F-400 (GAC) Extract of
the Aqueous Chlorination of Vanillic Acid.
15









The electrophilic addition of chlorine from hypochlorous acid to the
aromatic ring is favored by the presence of electron donating substituents
-OH and -OCH3. Methyl, monochloro and methyl, dimethoxy benzoate (6, 9)
were products formed by the substitution reaction. Monochloro and dichloro
methoxy phenol (1, 3) were produced by the oxidative decarboxylation of the
aromatic ring and the substitution of chlorine in the place of the electron
withdrawing substituent -COOH.

With the exception of methyl, dichloro dimethoxy benzoate and the
high molecular weight compounds, all products identified in this reaction
were exactly identical to those described by Larson and Rockwell (1979).
The high molecular weight species may have resulted from a free radical
reaction of the adsorbed species. Voudrias et al. (1982) observed high
molecular weight products when 2.6 dimethyl phenol was reacted with chlorine
on carbon. Apparently chlorine was not necessary for these reactions to
occur. Methyl, dichloro dimethoxy benzoate (9) was present only in trace
quantities and may have gone undetected in previous work. A reaction scheme
for the chlorination of vanillic acid in aqueous solution is presented in
Figure 9.


3.2 Chlorine/Adsorbed Vanillic Acid Reaction

The catalytic effect that activated carbon might have on the vanillic
acid/chlorine reaction was studied using a column of F-400 granular activated
carbon. Table 6 shows the experimental parameters for the column run.

The Soxhlet extract of the carbon produced the total ion current
chromatogram shown in Figure 10. Tentatively identified compounds found
in the extract are listed in Table 7 with their percent areas.

A series of substituted benzene compounds was identified. Trace
amounts of monochloro dimethoxy benzene (4) and dichloro methoxy phenol
were found. Trimethoxy (5) and tetramethoxy benzene (7, 13) were also
identified in the extract. Their monochloro derivatives, monochloro
trimethoxy benzene (6), and monochloro tetramethoxy benzene (12) were
present in small concentrations. The major compound present in this group
was tetramethoxy benzene with a 5.2% area. The molecular weights of
trimethoxy benzene (5), tetramethoxy benzene (7, 13) and their monochloro
derivatives (6, 12) are identical to the molecular weights of a series of
quinones. These were dimethoxy quinone (5), trimethoxy quinone (7, 13),
monochloro dimethoxy quinone (6), and monochloro trimethoxy quinone (12).
For example, dimethoxy quinone has the same molecular weight as trimethoxy
benzene (168) as well as a similar fragmentation pattern. McCreary et al.
(1982) showed evidence of both quinones and hydroxy benzenes formed in
this reaction.

Methylated carboxylic acids (methyl benzoates) were the most common
products formed in the reaction. Among these were methyl, methoxy hydroxy
benzoate (9), methyl, dimethoxy benzoate (10) (fully methylated vanillic acid)
















Table 5. Compounds Tentatively Identified in the Methylated Extract of the
Aqueous Chlorination of Vanillic Acid and % Area in the Chromatogram



MW % Area*

1. Monochloro methoxy phenol 158 1.0

2. Dichloro dimethoxy benzene 206 0.8

3. Dichloro methoxy phenol 192 0.9

4. Methyl, methoxy hydroxy benzoate. 182 55.5

5. Methyl, dimethoxy benzoate 196 7.8

6. Methyl, monochloro dimethoxy benzoate 230 27.0

7. Contaminant 192 2.6

8. D10 Anthracene (1.5) 188 2.9

9. Methyl, dichloro dimethoxy benzoate 264 traces**

10. Unknown (1 chlorine) 352 traces

11. Unknown (1 chlorine) 338 traces

12. Unknown (2 chlorines) 386 traces


* % areas were based on the area of the total ion current for each peak
divided by the sum of the ion current for all peaks.

** Less than 0.5%



















COOH


O OCH3 HOc
OH
S(4,5)

SHOC1, -CO2


Cl


Cl

O OCH3
OH
(1)


COOH
Cl (Cl
S OCHO3 'HOC1
Q JOCH
OH (6)

HOC1, -CO2

-Cl
C1

Cl OCH3
OH
(2,3)


COOH


2 OCH3
OH
(9)


*methylated derivatives were actually found




Figure 9. Reaction Scheme for the Chlorination of Vanillic Acid in
Aqueous Solution






















Table 6. Experimental Parameters for the Vanillic Acid/Chlorine/Carbon
Reaction


Column diameter, cm 1.9

Column length, cm 38.

Influent Vanillic Acid Conc., mg/l 2.6

Influent Chlorine, mg/l 10.0

Influent pH 6.0

Flowrate, 1/m2-min 81.1

Reaction time, min 2.0

Reaction volume, ml 45.8

Total volume passed through bed, 1 140.












20 25


Figure 10.


Total Ion Current Chromatogram of an F-400 (GAC) Extract of the
Vanillic Acid/Chlorine/Carbon Reaction.


it il~
,r~,,,~L~j ~,,,!








Table 7. Compounds Tentatively Identified in the Methylated Vanillic
Acid/Chlorine/Carbon Reaction and % Areas in the Chromatogram


MW1 % Area2

1. Unknown (non-chlorinated) 146 trace*

2. Monochloro dihydroxy methoxy benzene 174 trace

3. Unknown (non-chlorinated) 174 3.5

4. Monochloro dimethoxy benzene 172 trace

5. Trimethoxy benzene, or dimethoxy 168 trace
quinone

6. Monochloro trimethoxy benzene, or 202 2.6
Monochloro dimethoxy quinone

7. Tetramethoxy benzene, or 198 5.2
Trimethoxy quinone

8. Dichloro methoxy phenol 192 trace

9. Methyl, hydroxy methoxy benzoate 182 1.8

10. Methyl, dimethoxy benzoate 196 15.4

11. Unknown (non-chlorinated) 204 trace

12. Monochloro tetramethoxy benzene, or 232 trace
Monochloro trimethoxy quinone

13. Tetramethoxy benzene, or 198 trace
Trimethoxy quinone

14. Methyl, trimethoxy benzoate 226 trace

15. Methyl, monochloro dimethoxy benzoate 230 3.9

16. Methyl, trimethoxy benzoate 226 trace

17. Methyl, dimethoxy hydroxy benzoate 212 trace

18. Methyl, trimethoxy benzoate 226 trace

19. Methyl, monochloro trimethoxy benzoate 260 trace

20. Methyl, trimethoxy benzoate 226 30.1









MW1 % Area2


21. Methyl, monochloro trimethoxy benzoate 260 2.0

22. Methyl, trimethoxy benzoate 226 trace

23. Methyl, monochloro trimethoxy benzoate 260 trace

24. Methyl, monochloro trimethoxy benzoate 260 trace

25. Methyl, monochloro trimethoxy benzoate 260 trace

26. Dimethoxy benzene dicarboxylic acid- 254 trace
dimethyl ester

27. Monochloro methoxy hydroxy benzene 274 trace
dicarboxylic acid dimethyl ester

28. Dimethoxy hydroxy benzene dicarboxylic 270 trace
acid dimethyl ester

29. Methoxy hydroxy benzene tricarboxylic 298 trace
acid trimethyl ester


* Trace represents less than 1%.

1 The molecular weights in the Table are the weights of the partially
or fully methylated compounds found in the extract.
2 % Areas were based on the area of the total ion current for all peaks.








and its chloro derivative (15). Five isomers of methyl, trimethoxy
benzoate (14, 16, 18, 20, 22) and five isomers of the monochloro
derivative (19, 21, 23, 24, 25) were identified in the extract. One
additional compound, methyl, dimethoxy hydroxy benzoate (17) was present
in trace quantities. The major compounds in this group of methyl
benzoates were an isomer of methyl, trimethoxy benzoate (20) and an
isomer of methyl monochloro trimethoxy benzoate (25).

One last group identified in the reaction was poly carboxylic
acids. This group included dimethoxy benzene dicarboxylic acid -
dimethyl ester (26) and its monochloro derivative (27). Dimethoxy
hydroxy benzene dicarboxylic acid dimethyl ester (28) and methoxy
hydroxy benzene tricarboxylic acid trimethyl ester (29) were also
among this group. Only trace amounts of these compounds were detected.

Methyl, dichloro dimethoxy benzoate (dichloro vanillic acid),
one of the compounds identified in the influent to the column, was not
found in the activated carbon extract. It was only found in trace
quantities in the blank reaction and was possibly below the detection
limit in this reaction. All four other reaction products (Figure 9)
were found. A major component not present in the influent to the column,
but present in the carbon extract, was a monochloro trimethoxy benzene
or monochloro dimethoxy quinone (6) (MW 202). Its structure was confirmed
by the use of the EPA/NIH Mass Spectral Data Base (1978).

Five isomers of methyl, trimethoxy benzoate were found (14, 16,
18, 20, 22). It is not apparent how five isomers of methyl, trimethoxy
benzoate could be formed from vanillic acid. The three expected isomers
of methyl, trimethoxy benzoate are shown in Figure 11. Perhaps the
fourth and fifth isomers are derived from an impurity in the reaction
mixture. In addition there is the possibility of loss of -OH or -OCH3
from the vanillic acid molecule followed by hydroxylation at another
site. Mass reconstruction was done to verify the presence of five
compounds with a molecular weight of 226. Figure 12 shows the mass
reconstructed chromatogram for the molecular ion of 226. Three major
peaks (b,c,d) with abundances of 4.0%, 4.3% and 91.0% respectively
were found. These could correspond to the expected isomers of methyl,
trimethoxy benzoate. Peaks (a) and (e) were present in abundances of
less than 0.4%. The retention times of actual standards would have to
be compared to this extract in order to determine the correct isomer.
Five isomers of methyl, monochloro trimethoxy benzoate (19, 21, 23, 24, 25)
were detected in the extract. The five monochloro isomers of methyl,
trimethoxy benzoate are consistent with the substitution pattern of one
molecule of chlorine into each of the three structures (I,II,III)
(Figure 11) as shown in Figure 13. The five monochloro isomers were
found using a mass reconstructed chromatogram at a mass intensity of
260, the molecular ion of these species. Figure 14 shows the mass
reconstructed chromatogram. Five major peaks (a,b,c,d,e) with abundances
of 0.5%, 2.9%, 2.7%, 2.1%, and 91.7% respectively were present.
Another small peak (f) with an abundance of less than 0.2% was found
in the reconstruction. This sixth isomer could be due to the loss of
-OH or -OCH3 from the vanillic acid molecule followed by hydroxylation
at another site.
























x (?) -


COOH
OH OCH3


OH

I


COOH


S ---- Y (?)
. OCH3
OH


COOH


OH O OCH3
OH


COOH
OHH
OCH3
OH


*methylated derivatives were actually found




Figure 11. Methyl, Trimethoxy Benzoate Isomers


a b c


Molecular Ion 226


Total Ion Current Chromatogram



1~t-


Figure 12. Mass Reconstructed Chromatogram for the Molecular Ion of 226





















COOH COOH

OH OCH OH J] OCH3

OH OH
I II






COOH COOH COOH
OH c OH Cl Cl

Cl' OCH3 OCH3 OH OCH3
OH OH OH

*methylated derivatives were actually


COOH
OH

00 OCH3
OH
III






COOH COOH
,O OH Clf OH

ClZ OCH3 OCH3
OH OH

found


Figure 13. Methyl, Monochloro Trimethoxy Benzoate Isomers.















e


a b c


Molecular Idn 260

f


._-~-_-..~-


Figure 14. Mass Reconstructed Chromatogram for the Molecular Ion 260.








Compound


OCH3 COOCH3
OC3 30.1%
OCH3
OCH3
(20)

COOCH
OCH 3
a31 OCH26.4%
C1 OCH3
OCH3
(25)

COOCH3


OCH3 15.4%
OCH3

(10)

0

(OCH3 4 (OCH3)3 5.2%


(7)


Cl COOCH3
33.9%

OCH3 3.9%
OCH
3
(15.)



Figure 15. Major Products Formed in the Vanillic Acid/Chlorine/Carbon
Reaction


% Area






COOH COOH


Cl


OH OCH
OH
(19,21,23,
Cl 24,25)
Cl


O CH3
OH (8)
Cl


(OH) 2


OCH 3
OH OH
(14,16,18,20
Cl 22)


(4CH
OH (4)


(OH)


OH (12)
(12)


2,

N 3CH
OH (7,13)


NOCH 3
OH
(9,


COOH


O
I O t (2


10)


C COOH
C3


SOCH3
OH (15)


N OCH3 OH- OCH3
OH OH /* 3
OH OH (2,6)
(5)


O


(COOH)2

CH3
8)

COOH

CH3

7)

(COOH)3


.*methylated derivatives were
actually found


Figure 16.


Reaction Scheme for
the Vanillic Acid/
Chlorine/Carbon
Reaction.


0


OCH3
0
J o


O O
(OH Cl

OCH


S(7,13) l
OCH 3 0
0 (7,13) 0


1 OCH3
OH (28)
(28)


OH



(6)

(OH)2


(12)


COOH COOH


COOH








Compounds 1, 3 and 11 (Table 7) could not be identified. Compound
1 had a molecular weight of 146 and was not chlorinated. Compound 3 had
a molecular weight of 174 with a fragmentation pattern of m/e 159, 143,
115, and 69. Its area was 3.5% of the cumulative area of all the products
formed. Finally, compound 11, a non-chlorinated aromatic compound, had
a molecular weight of 204.

The major products formed in the reaction are shown in Figure 15.
Methyl, trimethoxy benzoate (20) and its monochloro derivative (25)
represented more than 50% of the area occupied by all compounds. Methylated
vanillic acid (10) and its monochloro derivative (15) occupied an area of
15.4% and 3.9% respectively. Compound 7, a tetramethoxy benzene or
trimethoxy quinone, had an area of 5.2%.

A general reaction scheme for the vanillic acid/chlorine/carbon
reaction is shown in Figure 16.


3.3. Aqueous Chlorination of Ferulic Acid

The chlorination products of ferulic acid were determined using a
blank reactor column. The experimental parameters, similar to those
used in the chlorination of vanillic acid are presented in Table 8.

The effluent of the column was reduced with sodium sulfite and
pumped through a 3-inch bed of F-400 granular activated carbon. The
carbon was then removed and Soxhlet extracted to give the total ion
current chromatogram shown in Figure 17. This chromatogram is much more
complex than the one shown in Figure 8 for the aqueous chlorination of
vanillic acid under the same conditions.

A list of the compounds tentatively identified in the aqueous
extract and percent area in the chromatogram is presented in Table 9.
It is obvious that the aqueous chlorination of ferulic acid yielded
many more products than the vanillic acid chlorination.

Phenol (3) and its monochloro, dichloro, and trichloro derivatives
(5, 8, 9) were present in the extract in trace quantities. Phenol may
have been present as a contaminant in either the water or the ferulic
acid used in the experiment. Monochloro trimethoxy phenol (12) and
dichloro dimethoxy benzene (10), chlorination products of vanillic acid,
were also found.

A series of carboxylic acids were extracted from the activated
carbon. Methyl benzoate (4), methyl, dimethoxy benzoate (16), methyl,
monochloro and dichloro dimethoxy benzoate (18, 19) and methyl,
tetramethoxy benzoate (24) were among the methylated carboxylic acids
identified. Two isomers of methyl, monochloro methoxy benzoate (13, 14)
were present at 2.8% and 2.7% of the total area in the chromatogram.
Methyl, dichloro dimethoxy benzoate was present at an abundance of 2%.






















Table 8. Experimental Parameters for the Ferulic Acid/ Chlorine
Blank Reactor Column


Column diameter, cm 1.9

Column length, cm 35.

Influent ferulic acid conc., mg/l 2.6

Influent chlorine conc., mg/l 10.0

Influent pH 6.0

Flowrate, l/m2-min 69.1

Reaction time, min 2.0

Reaction volume, ml 38.6

Total volume passed through bed, 1 18.8

























, il 3 1 b i 3
UiIJ U0 I_,1 39 40




Figure 17. Total Ion Current Chromatogram of an F-400 (GAC) Extract of the Aqueous
Chlorination of Ferulic Acid









Table 9. Compounds Tentatively Identified in the Methylated Extract
of the Aqueous Chlorination of Ferulic Acid and % Area in
The Chromatogram.


MW1 % Area2


1. Unknown (non-chlorinated) 126 trace*

2. Acetophenone 120 1.1

3. Phenol 94 trace

4. Methyl benzoate 136 0.3

5. Chloro methoxy benzoate 142 0.6

6. Unknown (1 chlorine) 150 1.5

7. Chloro-phenol 128 trace

8. Dichloro methoxy benzene 176 trace

9. Trichloro methoxy benzene 210 trace

10. Dichloro dimethoxy benzene 206 1.0

11. Monochloro trimethoxy phenol 218 0.5

12. Dimethoxy benzaldehyde 166 29.0

13. Methyl, monochloro methoxy benzoate 200 2.8

14. Methyl, monochloro methoxy benzoate 200 2.7

15. Monochloro methoxy acetophenone 184 1.5

16. Methyl, dimethoxy benzoate 196 6.1

17. Monochloro dimethoxy acetophenone 214 10.0

18. Methyl, monochloro dimethoxy benzoate 230 0.8

19. Methyl, dichloro dimethoxy benzoate 264 2.0

20. Methyl, hydroxy methoxy cinnamate 208 4.3

21. Trimethoxy B-chloro styrene 228 1.8









Table 9., con't.


MW1 % Area2


22. Trimethoxy B-chloro styrene 228 1.8

23. Methyl, monochloro hydroxy 242 1.8
methoxy cinnamate

24. Methyl, tetramethoxy benzoate 256 0.9

25. Trimethoxy B-chlorostyrene 228 1.8

26. D10 Anthracene (I.S.) 188 1.3

27. Tetramethoxy B-chlorostyrene 258 0.5

28. Methyl, dimethoxy cinnamate 222 2.2

29. Unknown 236 2.3

30. Methyl, monochloro dimethoxy cinnamate 256 3.9

31. Tetramethoxy B-chlorostyrene 258 3.8

32. Methyl, monochloro hydroxy 242 4.2
methoxy cinnamate

33. Methyl, monochloro dimethoxy cinnamate 256 1.0

34. Unknown 270 1.3

35. Methyl, chloro trimethoxy cinnamate 286 1.0

36. Methyl, chloro dimethoxy hydroxy 272 trace
cinnamate

37. Unknown 380 trace

38. Unknown (1 chlorine) 380 trace

39. Unknown (1 chlorine) 428 trace









Table 9., con't.


MW1 % Area2

40. Unknown (1 chlorine) 428 trace

41. Unknown (2 chlorines) 408 trace



* Represents less than 0.5%

SMolecular weights are the weights of the partially or fully methylated
compounds
% Areas are based on the area of the total ion current for each peak
divided by the sum of the ion currents for all peaks.









The major compound in this group was methyl, dimethoxy benzoate (16),
present in 6.1% area in the chromatogram.

The presence of dimethoxy benzaldehyde was confirmed by comparison
of its mass spectrum to published spectra. This represented the major
compound found in the aqueous chlorination reaction products collected
on carbon.

Another group of products in the extract were the acetophenones.
Among these were acetophenone (2), monochloro methoxy acetophenone and
monochloro dimethoxy acetophenone (17). Monochloro dimethoxy acetophenone
(17) was the most abundant compound in this group with a 10% area in the
chromatogram.

Especially interesting was the lack of dimethoxy B-chlorostyrene
and chloro dimethoxy B-chlorostyrene in the extract. These were previously
reported aqueous chlorination products of ferulic acid by Norwood et al.
(1980). Trimethoxy and tetramethoxy B-chlorostyrenes were found in the
extract, apparently formed as a result of hydroxylation of the aromatic
ring. Hydroxylation has previously been shown to be a common reaction
for the reaction of organic with chlorine on carbon. Apparently the
chlorine influent was not required for this reaction to occur when the
products were concentrated.

The methyl derivative of ferulic acid (28) and two isomers of its
fully methylated monochloro derivative (30, 33) were detected in the
reaction extract. Methyl, chloro trimethoxy cinnamate (35); methyl,
chloro dimethoxy hydroxy cinnamate (36) and methyl, chloro hydroxy
methoxy cinnamate (23, 32) were also present in the extract. Unreacted
ferulic acid (20, 28) was found in an abundance of 6.5%. Monochlorinated
ferulic acid was present in an abundance of 10.9%

Several of the mass spectra of the high molecular weight unknowns
(27, 28, 39, 40, 41) are discussed in Appendix A.

The major products identified in the ferulic acid aqueous chlorination
reaction are listed in Figure 18. Major compounds included dimethoxy
benzaldehyde (12), monochloro ferulic acid (23, 32, 33) monochloro dimethoxy
acetophenone (17), and methyl, dimethoxy banzoate (16). These compounds
were apparently produced by the oxidative decarboxylation of the unsaturated
alkyl chain of ferulic acid. Other major products included unreacted
ferulic acid (20, 28) at 6.5%, methyl chloro methoxy benzoate (13, 14)
at 5.5% and tetramethoxy B-chlorostyrene at 4.3% of the total area.

The chlorination of ferulic acid was studied by Norwood et al.
(1980). They reported a series of chlorinated substitution products and
chlorophenols at a 0.5 chlorine/carbon molar ratio which broke down upon
further reaction to chloroacetic acids at a 2.0 chlorine/carbon ratio.
Only monochloro ferulic acid, reported by Norwood et al. (1980) as a
chlorination product, was detected in appreciable amounts in this experiment.
The rest of the compounds reported by Norwood were not present in detectable












% Area Compound


0
C-OCH3



(- OCH3

OCH3


OCH3
0
II
CH= CH- C OCH3

Cl,3

OCH'
3
OCH3


10.9%


10.0%


CH= CHC1



C (OCH3)4


OCH3


OCH3


OCH3


6.5%


Figure 18.


Major Products Formed in the Ferulic Acid/Chlorine Reaction
Collected on Carbon


CHO


6.1%


OCH3


5.5%


4.3%


OCH3


OCH3


% Area


Compound









amounts. Major compounds reported in this experiment were aldehydes,
acetophenones and carboxylic acids, none of which were found in the work
of Norwood et al. (1980). The chlorostyrenes were found on the carbon
in this experiment as the hydroxylated isomers. Presumably the adsorbed
compounds were acquiring -OH from the carbon surface with incorporation
into the aromatic ring. The chlorine to carbon ratio used in our studies
(1.05) would indicate that the major chlorinated products should be mono
and dichloro ferulic acid, unchlorinated ferulic acid, methoxy hydroxy
B-chlorostyrene, chloro methoxy hydroxy B-chlorostyrene and chloroform
according to Norwood et al. Although chloroform would not be detected
by our procedures, we found many additional products as previously
described. This may be resolved by noting differences in experimental
design between our reactions and those of Norwood. A substantial
difference in the experimental procedure was that the organic in the
reduced aqueous solution were concentrated by adsorption on carbon
followed by Soxhlet extraction. This was done to duplicate as closely
as possible the carbon/chlorine experiment. It is apparent that
reactions of organic may take place on the carbon surface without
the presence of chlorine oxident. Chen et al. (1983) similarly observed
that BHT butylatedd hydroxy toluene) was converted to an aldehyde on
carbon without any oxidant (C102 or C12) in the influent. The reactions
were attributed to the presence of free radicals on the carbon surface.
Free radicals have been directly demonstrated by electron paramagnetic
resonance in carbonaceous materials (Lewis and Singer, 1981).

The aqueous chlorination of ferulic acid was repeated in order
to resolve conflicts between observed results and those reported by
Norwood et al. The types of products observed from the chlorination of
ferulic acid in the previous experiment was far greater than those
anticipated based on similar work by other researchers. A liquid-liquid
extraction technique was employed to concentrate the organic produced
in the aqueous chlorination reactions. This was done to reduce the
possibility of carbon catalyzed reactions taking place during concentration.
In order to determine only those products formed from aqueous chlorination
of ferulic acid, approximately 5 liters of reactor column effluent were
collected, reduced with excess sodium sulfite (Na2SO3) and stirred for
12 hours with 100 milliliters of methylene chloride (MeCIl). Experimental
parameters for the second ferulic acid chlorination experiment are given
in Table 10. The total ion current chromatogram from the aqueous extract
is shown in Figure 19.

The liquid-liquid extraction technique produced major differences
from those found using a carbon column to trap the reduced effluent.
Dichloro, dimethoxy benzene (10) and monochloro methoxy acetophenone (15)
were the only reaction products common to both extraction techniques.
Dimethoxy B-chlorostyrene, and two isomers of chloro dimethoxy
B-chlorostyrene were found, consistent with previous observations by
Norwood et al. The major product formed in the aqueous chlorination
of ferulic acid was chloro dimethoxy B-chlorostyrene. A list of the
compounds tentatively identified in the aqueous extract and percent
areas in the chromatogram are presented in Figure 20.






















Table 10. Experimental Parameters for the Aqueous Chlorination of
Ferulic Acid


Column Diameter, cm 0.9

Column length, cm 29.7

Influent Ferulic Acid Conc., mg/l 3.35

Influent Chlorine Conc., mg/l 10.2

Influent pH 6.85

Flowrate, ml/min 11.50

Reaction time, min 1.64

Reaction volume, ml 18.9

Total volume passed through the bed, 1 5.

Type of water used tap


























** SPECTRUHI DISPLAY/EDIT -
CIHN1HCL ELANK RUII PH-6i2UL+IS
ll.'2/82kR*B+*30-28Oe5/IIN


FRII 9131 .* SPECTRUM DISPLAY/EDIT *
1ST SC,'PCG 1RIIK RUN PH=6*2UL-IS
X- .50 Y- 1.00RB**30-28095/IHN


FRN 9131 ** SPECTRUM
1ST SC/PGi 414 BLANK RUN PHa6**2UL+IS
X- .S5 Y- 1.00**KRB**3O-280@5/MIH


1


1
6








LJI


Figure 19.


Total Ion Chromatogram for the Methylene Chloride Extract of the Aqueous
Chlorination of Ferulic Acid'


4* FRII 9131
1ST SC/PCG 849
X- .s5 Y. i.00


7


29 38 31 37 3 39 4


4 44 4 46 4 4 4 5 5


I


--


use Amenst


.. ----~--~-;r---r~


34 4 4 4


T "J


' '
5 5 ? 2%


* a ^
















CH= CHC1


CH= CHC1


- C1
OCR3


OCH3


OCH3


(2) (Cl)2

OCH3
OCH3
0
II
C CH3


(3) 3C1


OCH3


OCH3


Cl1

OC3 3
OCH3


.Cl


OCH3


Figure 20. Compounds Tentatively Identified in the Aqueous Chlorination
of Ferulic Acid.


OCH3









Although the chlorinated styrenes were found in large quantities
in the aqueous extract of chlorinated ferulic acid, these products
were not found in the aqueous extract collected on carbon. These
compounds were apparently hydroxylated via a free radical mechanism
at the carbon surface. Trimethoxy and tetramethoxy B-chlorostyrenes
were recovered off the carbon (Table 9). It is significant that free
chlorine residual was not required for this reaction to occur. It
was determined from these two experiments that a liquid-liquid extraction
technique provided more accurate results for the reaction products
of ferulic acid and aqueous chlorine in the absence of carbon.

The most common reactions in the aqueous chlorination of ferulic
acid were oxidative decarboxylation of the alkyl side chain with
concomitant substitution of chlorine, oxidation of the benzylic carbon
to form acetophenones, and substitution of chlorine into the aromatic
ring.


3.4. Ferulic Acid/Chlorine/Carbon Reaction

The catalytic effect of activated carbon on the reaction of
ferulic acid and chlorine was studied using F-400 GAC. The experimental
parameters are listed in Table 11.

A total ion current chromatogram (electron-impact mode) of the
methylated extract of the activated carbon used in the experiment is
shown in Figure 21. The extract was also run in the Negative Ion -
Chemical Ionization mode (NI/CI) as an aid in determining which compounds
were chlorinated. Negative Ion Chemical Ionization also enabled
the molecular weight to be determined for those chlorinated compounds
which had a weak molecular ion in the electron impact mode. The NI/CI
chromatogram is shown in Figure 22. Table 12 presents the tentatively
identified compounds in the extract of the ferulic acid/chlorine/carbon
reaction. The compounds are listed in the order of elution in the
total ion current chromatogram (Figure 21) with percent areas in the
chromatogram.

Most of the compounds tentatively identified in this reaction
were present in the influent to the carbon column. A major series
of products; monochloro compounds of molecular weight 204 and the
dichloro derivatives of molecular weight 238 were intriguing. A
logical structure could not be derived from the fragmentation pattern
of the low resolution mass spectra. The electron impact spectra
demonstrated a loss of -C1 or -OCH3 to yield a large M-35 and M-31
fragment ion. The molecular ion could be confirmed only through the
use of NI/CI spectra (Figure 23). The spectra of the MW 238 compound
was similar, with a large M-35 ion and another fragment corresponding
to a loss of -OCH3. Although these compounds comprised approximately
30% of the total area, they were not present in the influent to the
column and were not previously reported chlorination products of























Table 11. Experimental Parameters for the Ferulic Acid/Chlorine/Carbon
Reaction


Column diameter, cm

Column length, cm

Influent Ferulic Acid Conc., mg/l

Influent Chlorine Conc., mg/l

Influent pH

Flowrate, l/m2-min

Reaction time, min

Reaction volume, ml

Total volume passed through bed, 1

Type of water used


1.9

38

2.6

10.0

6.0

77.6

2.1

45.8

132.

distilled
































-P
N)


Figure 21.


Total Ion Current Chromatogram of an F-400 (GAC) Extract of Ferulic Acid/Chlorine/Carbon
Reaction






























Figure 22.


I I __ .


Negative Ion Chemical Ionization Chromatogram of an F-400 (GAC) Extract of the
Ferulic Acid/Chlorine/Carbon Reaction'


jir









Table 12. Compounds Tentatively Identified in the Methylated Extract
of the Ferulic Acid/Chlorine/Carbon Reaction and % Area in
the Chromatogram.

MW1 % Area2


1. Unknown (monochlorinated) 178 trace*

2. Unknown (monochlorinated) 178 2.5

3. Unknown (monochlorinated) 178 0.7

4. Unknown (monochlorinated) 204 0.8

5. Unknown (monochlorinated) 204 18.6

6. Unknown (monochlorinated) 204 0.6

7. Unknown (monochlorinated) 204 trace

8. Unknown (monochlorinated) 204 6.6

9. Monochloro trimethoxy phenol 218 trace

10. Unknown (dichlorinated). 194 2.3

11. Unknown (dichlorinated) 238 trace

12. Unknown (dichlorinated) 238 7.7

13. Methyl, methoxy hydroxy benzoate 182 0.8

14. Monochloro methoxy acetophenone 184 trace

15. Methyl, monochloro methoxy benzoate 200 3.0

16. Monochloro methoxy acotophenone 184 2.8

17. Unknown (monochlorinated) N.D. trace

18. Unknown (monochlorinated) N.D. 1.0

19. Unknown (dichlorinated) N.D. 1.0

20. Unknown (unchlorinated) 212 1.2

21. Methyl, monochloro dimethoxy 230 0.4
benzoate









Table 12., con't.


MW1 % Area2


22. Unknown (not chlorinated) N.D. 4.6

23. Dimethoxy B-chloro styrene 198 28.1

24. Contaminent from water 192 10.3

25. Dichloro methoxy acetophenone 218 trace

26. Methyl, tetra methoxy benzoate 256 trace

27. Unknown (not chlorinated) N.D. 5.7

28. Unknown (not chlorinated) 260 1.2




* Represents an area of less than 0.4%

1 Fully or partially methylated molecular weight of the compounds.

2 % Areas are based on the area of the total ion current for each
peak divided by the sum of the ion currents for all peaks.


/









ferulic acid (Norwood et al., 1980). In order to determine that these
products were not impurities in the reactants, the ferulic acid was
recrystallized from methanol-water and rerun in an identical experiment.
Products of molecular weight 204 and 238 with similar retention times
and spectra were detected again in high yield. The carbon extract
containing the unknown MW 204 compound was fractionated on a silica
gel column by elution with solvents of increasing polarity. Approximately
one milliliter of a methylated extract was eluted with benzene, ether,
methylene chloride, methylene chloride methanol (1:1) and methanol.
The 204 MW compound was found at the greatest concentration in the
benzene eluate. This fraction was submitted to Dr. Roy King, UF
Chemistry Department, for high resolution mass spectral characterization
using a Kratos MS 30 double-focusing GC/MS instrument. The results
of this analysis gave the chemical formulas for the major fragments
of the compounds as shown in Table 13.

Since the molecular weight of the compound was 204, the molecular
formula could be interpreted as C8H904C1. Additional work demonstrated
that an unmethylated extract had no MW 204 peak and additional methylation
resolved the 204 peak into a MW 218 peak which had a similar fragmentation
pattern (see Figure 24).

A suggested formula for this compound is shown in Figure 25.
This is consistent with the structural formula and at least two reactive
hydrogens for the parent compound. This would also be consistent
with the variety of isomers found in the extract.

This polyhydroxy benzene structure for reactions of phenolic
acids on carbon is consistent with previous results for the carbon/
chlorine reaction of vanillic acid (McCreary et al., 1982) and the
chlorite/vanillic acid reaction (Voudrias et al., 1982).

A series of methyl derivatives of carboxylic acid were found.
Among these were methyl, methoxy hydroxy benzoate (13), methyl, monochloro
dimethoxy benzoate (21), methyl, monochloro methoxy benzoate (15), and
methyl, tetramethoxy benzoate (26). All the mass spectra of these
compounds compared with the corresponding mass spectra of the compounds
found in the aqueous chlorination of ferulic acid collected on carbon.
Compound 15 was the most abundant of the carboxylic acid derivatives
formed. Methyl, monochloro methoxy benzoate was also found in the
aqueous reaction in appreciable amounts. Two isomers of monochloro
methoxy acetophenone (14, 16) and its dichloro derivative (25) were
identified in the extract. In this reaction, dichloro methoxy acetophenone
(25) was present in only trace quantities and it may have been below
detectable limits in the influent to the carbon column. Another product,
dimethoxy B-chlorostyrene (23) was identified in the extract. It was
the major identified product of the reaction on carbon with a 28.1%
area. This is consistent with the results of Norwood et al. (1980)
and the blank reactor effluent concentrated with methylene dichloride
extraction. Since the chlorostyrene was extracted from the carbon











LARGEST 4:
LAST 4:


204.2,100.0
204.2,100.0


206.2, 43.9
205.2, 7.5


205.2, 7.5 207.3, 2.1
206.2, 43.9 207.3, 2.1
PAGE 1 Y = 1.00


100

80

60

40

20

0
4~.. .. ....401~


.. ~ ~ ~ 2 240~. _~; ..... 26;g. 29.3 0I(; -7~-1-. 1 3- -I ..-


Figure 23. Negative Ion Chemical Ionization Spectrum of MW 204 Compound.


LARGST 4: 169.1,100.0
LAST 4: 177.0, .1


173.1, 1 .8.
185.1, .1


59.0, 17.1 50.9, 14.1
204.1, .4 206.1, .1
PAGE 1 Y = 1.00


1 0

:3



40$


100





401
-J
086

Ae


2'0


I) I ~ .II


4i 60 SO


Figure 24. Electron impact mass spectra of the Unknown Compound #5 in the Ferulic Acid/Carbon/
Chlorine Reaction.


47


80

60

40

20


- ........ ....








Table 13. Structural Formulas of Major Ions of MW 204 Unknown

C8H904 169.05
C7H603C1 173.0104
C7H902 141.0577
C2H302 59.0159
C4H3 51.0234













(OH)3 CH2 N CH2N2 N20
MW 2 OCH (OCH3)3
OCH3- Cl (H)2 Cl
MW 190 MW 204 / MW 218


Figure 25. Possible Structure of the MW 204 Compound









surface in this experiment it is not apparent why the aqueous extract
collected on carbon had the hydroxylated isomer. Perhaps excess
sulfite in the effluent contributed to the formation of catalytic
products. The major products tentatively identified in the reaction
of ferulic acid with chlorine on carbon are presented in Figure 26.

The chlorination of ferulic acid in the presence of carbon was
duplicated to facilitate identification of unknown compounds. The
experimental parameters are shown in Table 15. A total ion current
chromatogram of the methylated extract is shown in Figure 27 and
products are tentatively identified in Table 16.

Several monochloro and dichloro methoxy acetophenones were
identified. In addition, a large peak with a molecular weight of
140 and a peak with a molecular weight of 192 were observed. The peak
with a molecular weight of 140 had mass fragments of 125 and 97,
corresponding to losses of 15 (CH3) and 43 COCH3) respectively. A
possible product could be an acetyl methoxy furan. Voudrias et al.
(1982) observed a similar product in a reaction of vanillic acid and
chlorite which was tentatively ascribed to a carboxy methyl furan.
The product with a molecular weight of 192 was tentatively identified
as 3-(dimethoxy phenyl) 2-propenal. The mass fragmentation was
identical to the spectra listed in the EPA-NIH libraries. In addition,
the peaks corresponding to the polyhydroxy chlorinated benzenes (204,
238) were also present. It is interesting that products were not
identical in the two ferulic acid/chlorine/carbon runs. The later run
used tap water instead of distilled water; however, all other parameters
were similar. Apparently the reactions of ferulic acid with chlorine
on carbon involve complex chemistry; products on carbon include
hydroxylation of the aromatic ring and oxidation of the alkyl group
to an aldehyde, acid or ketone. The hydroxylation of the aromatic
ring was similar to the products observed when vanillic acid was
chlorinated on carbon; however, the oxidation of the benzylic position
(possibly by a free radical mechanism) led to a plethora of oxidized
products. Free radical mechanisms for the reaction of organic on
the carbon surface have been advanced by Chen et al. (1983). These
researchers observed a marked decrease in the catalytic activity of
activated carbon beds after surface radicals had been deactivated
with diazomethane.

It is also significant that many of the products found on carbon
were formed with and without the presence of chlorine residual.
Hydroxylation of the aromatic ring and other reactions, such as the
oxidation of the alkyl chain to an aldehyde, acid or acetophenone,
apparently require only the carbon surface. This is quite significant
since carbon beds often receive influent with little or no chlorine
residual. Chlorine is reduced rapidly at the carbon surface (Snoeyink
et al., 1981), and studies have demonstrated that continued chlorination
can reduce the adsorptive capacity of carbon (Snoeyink et al., 1974;
Yohe and Suffet, 1979). Thus it is advantageous to limit the quantity
of chlorine that enters an activated carbon contactor.






















Table 15. Experimental Parameters for the Ferulic Acid/Chlorine/
Carbon Reaction--Second Experiment


Column Diameter, cm 0.9

Column Length, cm 29.7

Influent Ferulic Acid Conc., mg/l 3.35

Influent Chlorine Conc., mg/l 10.2

Influent pH 6.85

Flowrate, ml/min 5.40

Reaction time, min 1.74

Reaction volume, ml 9.39

Total Volume passed through bed, 1 70.

Type of water used tap





















Compound % Area

CH=CHC1

OC0 31.4
CH3
OCH3
(23)
COOCH3
Cl OCH 3.3

t~iJOCH3

(13)

COCH3
3"3.1
C1 OCH3 3.1

(16)




Figure 26. Major Products Formed in the Ferulic Acid/Chlorine/
Carbon Reaction



























I.S.


Figure 27.


Total Ion Chromatogram for the Ferulic Acid/Chlorine/Carbon Reaction (Second Experiment).




















- 0
) C -CH3

0
II
CH= CH- CH




O OCH3

OCH3


0
1 I
C-H


OCH3


Table 17.


(OCH3)2


OCH3


OCH3


"'U 2

)2





C1

OCH3


0
C- OCH
3


OCH3


OCH3


Compounds Tentatively Identified in the Ferulic Acid/
Chlorine/Carbon Reaction (Second Experiment).


CH3(









3.5 Aqueous Chlorination of Ferulic Acid at pH = 10.


In order to study the aqueous reaction products of ferulic
acid and chlorine under normal water treatment plant conditions,
ferulic acid was chlorinated at a pH of 10 and the reaction products
were examined. At this pH, the predominant chlorine species is OC1-,
whereas the previous work had HOC1 as the predominant chlorine species.
Chlorination of ferulic acid at high pH might occur in a water treatment
plant during the softening process and evaluation of the effect of pH
on this reaction would allow-determination of the optimal point of
chlorination.

Hypochlorous acid is an electrophilic reagent at either the
chlorine or the oxygen atom. At a pH greater than approximately 8.0
(Snoeyink and Jenkins, 1980), chlorine residual exists primarily as
hypochlorite anion (OC1-). It has generally been observed that more
rapid addition of chlorine occurs at pH regions where HOC1 predominates.
Compounds formed through chlorination followed by base-catalyzed
hydrolysis may occur in higher yield at higher pH, however (Boyce
et al., 1983).

The chlorination products of ferulic acid at a pH of 10 were
determined using a blank reactor as in previous experiments. The
experimental parameters are shown in Table 17. Both influent solutions
to the column were adjusted to pH = 10 with NaOH pellets and buffered
with 0.001 M phosphate.

The effluent from the column was collected in a 40 liter carboy
and a liquid-liquid extraction was performed on a 5 liter sample.
The total ion current chromatogram is presented in Figure 28.

Tentatively identified compounds and their percent areas in the
chromatogram are shown in Table 18. Both ferulic acid and chloro
ferulic acid were found as products in this reaction. The presence
of these compounds was expected and agreed with findings by Norwood
et al. (1980) for chlorination at lower pH.

Several isomers of dimethoxy B-chlorostyrene and chloro dimethoxy
B-chlorostyrene were found in high yield. These two compounds were
also found by Norwood et al.

Acetophenones were another group of products formed in the aqueous
chlorination of ferulic acid at pH 10. Chloro methoxy acetophenone and
chloro dimethoxy acetophenone were both found in moderate yield in
this experiment. These compounds were found in all previous experiments.
Thus, pH appeared to exert a minimal influence on the formation of
these compounds.
























Table 17. Experimental Parameters for the Aqueous Chlorination of
Ferulic Acid at a pH of 10.0.


Column Diameter, cm 0.9

Column Length, cm 29.7

Influent Ferulic Acid Conc., mg/l 3.0

Influent Chlorine Conc., mg/l 10.55

Influent pH 9.95

Flowrate, ml/min 11.19

Reaction time, min 1.69

Reaction volume, ml 18.9

Total Volume Passed, 1 5.

Type of Water Used tap





































Figure 28.


Total Ion Chromatogram for the Aqueous Chlorination of Ferulic Acid at pH = 10.0


















CH- CHC1


0
C CH






OCH3


1,2


OCH3


OCH,


0

CH= CH-C- OCHC


9,10


C1

j -OCH3

OCH3


OCH3


CH = CHC1


OCH3


0
II
CH=CH- C OCH3
1


0
II







OCH3


OCH3


OCH3


Tentatively Identified Compounds From the Ferulic Acid/
Chlorine/Carbon Reaction at pH = 10.0


4,5,6


OCH3


Table 18.


^f









3.6 Ferulic Acid/Chlorine/Carbon Reaction at pH 10


The effect of pH on the ferulic acid/chlorine/carbon reaction
was examined by adjusting the pH of influent solutions to 10 and
buffering these solutions with 0.001 M phosphate. The experimental
parameters are presented in Table 19.

The total ion current chromatogram of the methylated extract
of the activated carbon used in the experiment is shown in Figure 29.
Table 20 presents tentatively identified compounds in the extract of
the ferulic acid/chlorine/carbon reaction. Products are listed according
to abundance.


3.7 Aqueous Chlorination of a Natural Water Sample

Previous experiments were conducted for the purpose of understanding
the types of reactions which occur during aqueous chlorination of
compounds commonly found in natural waters. Compounds were selected
to determine the types of reactions which occur at specific functional
groups in order to provide a basis for the analysis of chlorination
products of naturally occurring humic and fulvic acids. These compounds
are of interest due to their ubiquitous nature in surface and ground
waters. Humic and fulvic acids are too complex to characterize
individually and many researchers have used model precursor compounds
to investigate specific reaction chemistry. Carbon-catalyzed chlorination
products are especially important with humic and fulvic compounds as
these products can be expected to occur in water treatment plant processes.

A natural surface water was obtained which had passed through a
cypress dome wetland and had been characterized by previous researchers
(Dierberg et al., 1980). The particular water chosen was Hatchet Creek
in Gainesville, Florida. This water was diluted with distilled make-
up water to achieve a humic acid concentration of approximately 5 mg/l
derived from total organic carbon (TOC). The TOC of humic acid was
based on a correlation between color and TOC developed for this water
by Dierberg et al. (1980). The influent solutions were buffered with
0.001 M phosphate and reacted in a blank reactor column as in previous
experiments. The experimental parameters are listed in Tables 21 and 22.

The total ion current chromatogram shown in Figure 30 indicates
the presence of relatively few products. Among the chlorinated products
found were chloroform, dibromochloromethane, and bromoform. The
total ion chromatogram for the humic acid/ chlorine / carbon reaction
is shown in Figure 31. In this chromatogram, several compounds are
present which were not found in the aqueous chlorination of this
natural water. Several compounds identified were phthalates and fatty
acid methyl esters. A mass reconstruction was performed on masses 74
and 149 and is shown in Figure 32. The mass 149 is indicative of
phthalates and the mass 74 is indicative of acid esters. Due to the























Table 19. Experimental Parameters for the Reaction of Ferulic Acid
and Chlorine on Carbon at pH = 10.0


Column Diameter, cm 0.9

Column Length, cm 29.7

Influent Ferulic Acid Conc., mg/l 3.0

Influent Chlorine Conc., mg/l 10.55

Influent pH 9.95

Flowrate, ml/min 5.80

Reaction time, min 1.61

Reaction volume, ml 9.39

Total volume passed through bed, 1 60.

Type of water used tap


























.* SPECTRUM DlIPLAY.'EIf T -* FRIl 9119 .. SPECTRUM DISPLAO/.EDIT ** FRH 9119 ** SPECTRUIt DISPLAY/EDIT ** FRII 9119
CIII/CL CARiON EXT PACT-PHlO+*2L2UL TIAMFLE WITH IS(IOUL I-6THIC/PGI IRBOII EXTRACT-PH=I*2UL SUL SAHTCP 411RE XTCT-PH1 UL PLE ITH UL TCP ON EXTCT-PH= UL PLE WITH IS(IOUL IISTN C/PGCI 23
10/.22/82**RRB I**30-2@0/MIII WITH 20 MIN X- .SO VY 1.BWKRBrt30-Z80S5/IIII WITH 20 HIN X- .50 Y- 1 .OKRB**30-20Su5/III WITH 20 111H X .50 Y. 1.00








S44
53 7
5. 4 1 1 4 1


2 2' 2 A 2'9 3b 3'1 3 b 3 3 3' 4V 41 4h 44 4k '6 4 4b 4 b f


Figure 29. Total Ion Chromatogram for the Ferulic Acid/Chlorine/Carbon Reaction at pH = 10.




















OCH3

OCH CH
3
OCH3


0
C-CH3
i 3


0
II
CH




OCH3
OCH3
0

C- OCH



OCHH
OH


OCH3


CH= CHC1




CH30 'OCH3
OCH3


0
II
C OCH
3


OCH3


0
II
CH= CH -C-H




OUH3OCH3
OCH


Compounds Tentatively Identified in the Ferulic Acid/
Chlorine/Carbon Reaction at pH 10.


Table 20.























Table 21. Experimental Parameters for the Aqueous Chlorine/Humic Acid
Reaction


Column Diameter, cm

Column Length, cm

Influent Humic Acid Conc., mg/l

Influent Chlorine Conc., mg/l

Influent pH

Flow rate, ml/min

Reaction time, min

Reaction volume, ml

Total volume passed through reactor, 1

Type of water used


1.9

15.

2.5

9.8

6.1

12.76

2.9

37.6

5.

distilled

























:- FECtI.l.1 DISPLA',(lIT ** FRI 9488 ** SFPCTRUJI rtf.kPLAY/EDIT FN 4'8 SFECTRUI DI 'LAY'EZJIT ** Fd 94A8
SrFT;EL. rLAIIK FOP hflTCHETT CREEK RUN IST SC/PCG *I LAIIK FOR hATCHETT CREE.. RUII IST SC/PG, 435 ELIIK FOR HATCHETT CFEEK RUII IST SCPG: 369
S 20O AT 10 IHN AF1EPR S HOLD, THRESH= X= 360 Y- I.001 AT 10MIII AFTER 5 HOLI, THPESH- X- .60 Y. 1.001 AT IOlTIHN RFTER S HOLD, THRESH- X. .*0 Y I .00



















8 'i '70 11 1 1^ i I' | t 2 2r 14 231 24 aV 2-" 34 3 36 3 34 3' 4 4' 42 43 44


Figure 30. Total Ion Chromatogram for the Chlorine/Humic Acid Reaction























Table 22. Experimental Parameters for the Aqueous Humic Acid/Chlorine/
Carbon Reaction


Column Diameter, cm

Column Length, cm

Influent Humic Acid Conc., mg/l

Influent Chlorine Conc., mg/l

Influent pH

Flow rate, ml/min

Reaction Time, min

Reaction Volume, ml

Total volume passed through bed, 1

Type of water used


1.9

38.

2.5

9.8

6.1

25.5

1.80

45.8

90.

distilled

































tr SPECTPUK DISPLAY/EDIT FRN 9S26 ** SPECTRUM DISPLAVEDIT FRN 95 6 ** SPECTRUM DISPLAY/EDIT *FR 9626
HATCHET CREEKI/CL CAPPON EXT.*2UL+IS 1ST SC/PSI :REEK/CL CARBON EXT**ZUL.IS IST SC/PZs 433 CREEK/CL CARBO1 EXT.*2UL+IS IST SC/PGi 865
KRB-*3/19 83-30-2s8OS/I1IN X- .58 Y- 1.0.0 )/83*30-28085/IIN XY .SO Y- 1.00 9183r*30-280056/1-1 X. .50 Y. .00





















T12
A


Figure 31. Total Ion Chromatogram for the Humic Acid/Chlorine/Carbon Reaction









difficulty of identification of compounds present in this experiment,
the significance of the products was evaluated by submitting the
carbon extract for mutagenicity testing by the Ames assay (Ames and
Yanofsky, 1971; Ames et al., 1975; Ames, 1979).

The reversion of histidine-dependent Salmonella bacteria by
mutagenic compounds was determined in a two-day test using a variety
of altered bacterial strains. This test has been used on environmental
samples and as a result, low levels of direct-acting mutagens are
reported to be ubiquitous in U.S. water supplies as a result of
chlorination practice (Pelon et al., 1977; Loper et al., 1978; Glaz
et al., 1978). A preliminary experiment to test mutagenicity was
conducted on the natural water/chlorine/carbon extract and the results
are presented in Table 23 for TA-98 (frameshift mutagenicity) and
TA-100 (base pair mutagenicity).

It was obvious that more than a two-fold increase in mutagenicity
occurred over the spontaneous rate for the TA-100 carbon extract.
This indicated that mutagenicity, formed through the chlorination of
fulvic acid in solution or on carbon can be extracted from the carbon
surface and tested. Identification of products in future experiments
will, therefore, be limited to those extracts which exhibit mutagenicity.
Mutagenic fractions of carbon extracts will be isolated and then compared
to the mutagenicity present in chlorinated aqueous solutions.














Table 23. Mutagenicity of Carbon Extracts


Strain


Concentration


Carbon Samplea

TA-100

TA-98



Standard

TA-100

TA-98


# of Revertants


10 iI

10 V1


10 1I

10 il


328 339

26 17


200

9


251 225 234

231 217 194


aSpontaneous revertants: TA-98 = 11.4+-3

TA-100 = 71+-20

2-nitrofluorene






























.. ;FECTFUII DI TFLfEr.[LIT FNi -s PE(CTRUM DAIPL'/EBI *i
HATCHET CFEEK CL CaPF'OI E::T-ZIiUL*+I 1IT CFI;: I CFEEK,/L CRPEOII EXT*2UL+I,
PFl 3 1 -l3+. -2M94Ci.lll ;::= ,.50 y= .00 19 .3*, 30- :00l5. M'11


F- -- *- -rjIuJ Irr -,L


FII 95256 SPECi-TRUtl IIPL.H I Y/UI I *.l
1ST SC/F'Gs 44ET CREEK/CL CAPBOII E;T**2UL+IS
X= .50 Y= 1 .3/19/83ti30-28' 5/ 111


I- s J _4 __ ... A__ A A _


Figure 32. Mass Reconstructed Chromatograms for the Humic Acid/Chlorine/Carbon Reaction


1ST SCIPGI 883
Si .50 Y= 1 .0


race new


- -- --


-















CHAPTER 4, CONCLUSIONS


1. Chlorine will react with vanillic acid at the carbon surface
to produce a variety of hydroxylated aromatics. In addition, carbon
dioxide can be substituted from the surface of the carbon into the
aromatic ring.

2. Chlorine is not required for many of the carbon-catalyzed reactions.
A reduced solution of vanillic acid and chlorinated products collected
on activated carbon resulted in many high molecular weight chlorinated
compounds. Several of the compounds appeared similar to those found
by Voudrias et al. (1982) during the chlorination of phenols on carbon.

3. When ferulic acid and chlorinated products were reduced with
sodium sulfite and collected on carbon, a variety of products were
formed that were different than typical aqueous chlorination products.
These included hydroxylated chlorostyrenes, a benzaldehyde, and several
chlorinated acetophenones.

4. The aqueous chlorination products of ferulic acid were concentrated
by methylene chloride extraction. The major products isolated in
this case were those previously reported by other researchers (Norwood
et al., 1980).

5. The chlorination of ferulic acid on carbon was repeated in two
separate experiments. In the first experiment, the major products
were a series of tentatively identified tetrahydroxy benzenes. This
was determined through the use of negative ion chemical ionization
mass spectrometry, high resolution mass spectrometry and sequential
methylation with diazomethane. The second experiment found similar
compounds as well as a large peak for 3-(dimethoxy phenyl)- 2-propenal.
The products were not the same in both cases, possibly due to the use
of tap water in the later experiment.

6. The reaction of ferulic acid at pH 10 was also evaluated. The
major products found in the carbon reaction were trimethoxy benzene,
a chlorinated acetophenone, the previous propenal and vanillic acid.
These products were similar to those identified in previous carbon/
chlorine/adsorbed organic reactions.

7. A natural brown-colored water was chlorinated and applied to an
activated carbon bed. Preliminary analysis indicated that many









phthalates and fatty acid methyl esters were recovered from the
carbon surface. The carbon extract was tested in the Ames mutagenicity
assay and found to be positive in the TA-100 strain (base pair
reversion). Thus, mutagenic compounds can be extracted from the carbon
surface. The desorption of these compounds during normal carbon usage
is currently under evaluation.














REFERENCES


Bente, P.F., Walton, J.H. (1943). The Catalytic Activity of Activated
Nitrogenous Carbons. J. Phys. Chem., 47, 133.

Boyce, S.E., Hornig, J.F. (1983). Reaction Pathways of Trihalomethane
Formation From the Halogenation of Dihydroxyaromatic Model Compounds
for Humic Acid. Environ. Sci. Technol., 17, 262.

Chen, A.S.C., Larson, R.A., Snoeyink, V.L. (1982). Importance of
Surface Free Radicals in an Aqueous Chlorination Reaction (Indan-
C102) Promoted by Granular Activated Carbon. Submitted to Carbon.

Christman, R.F., Johnson, J.D., Pfaender, F.K., Liao, W.T., Norwood,
D.L., Alexander, H.L. (1978). Natural and Model Aquatic Humics:
Reactions with Chlorine. Chapter in Water Chlorination. Environ-
mental Impact and Health Effects. Vol 2. R.L. Jolley, H. Gorchev,
C.H. Hamilton, Eds. Ann Arbor Sci. Pub., Ann Arbor, MI., 15.

Chriswell, C.D., Ericson, R.L., Junk, G.A., Lee, K.W., Fritz, J.S.,
Svec, H.J. (1977). Comparison of Macroreticular Resins and
Activated Carbon as Sorbents. J. Am. Water Works Assoc., 69, 669.

Culp, R.L., Wesner, G.M., Culp, G.L. (1978). Handbook 6f Advanced
Wastewater Treatment. Van Nostrand Reinhold Company, New York, N.Y.

Eight Peak Index of Mass Spectra (1974). Mass Spectrometry Data Centre,
AWRE Reading, RG 7, 4 PR, U.K.

Farmer, W., Firth, J.B. (1924). The Catalytic Activity of Carbons
from Aromatic Hydrocarbons and Some Derivatives. J. Phys. Chem.,
28, 1136.

Garten, V.A., Weiss, D.E. (1957). Australian J. Chem., 10, 309.

Glatz, B.A., Chriswell, C.D., Arguello, M.D., Svec, H.J., Fritz, J.S.,
Grimm, S.M., and Thompson, M.A. (1978). Examination of Drinking
Water for Mutagenic Activity. J. Am Water Works Assoc., 70, 465.

Ishizaki, C., Cookson, Jr., J.T. (1974). Influence of Surface Oxides
on Adsorption and Catalysis with Activated Carbon, Chapter in
Chemistry of Water Supply, Treatment and Distribution., A.J. Rubin,
Ed., Ann Arbor Sci. Pub., Ann Arbor, MI., 201.









King, A. (1936). Studies in Chemisorption on Charcoal. Part VIII.
The Influence of Temperature of Activation of Charcoal on a)
Catalytic Oxidation of Salts b) Catalytic Decomposition of H202.
J. Chem. Soc. Part 2., 1688.

Larson, E.C., Walton, F.H. (1940). Activated Carbon as a Catalyst in
Certain Oxidation-Reduction Reactions. J. Phys. Chem., 44, 70.

Larson, R.A., Rockwell, A.L. (1977). Gas Chromatographic Identification
of Some Chlorinated Aromatic Acids, Chlorophenols, and Their
Aromatic Acid Precursors. J. of Chromatography, 139, 186.

Larson, R.A., Rockwell, A.L. (1979). Chloroform and Chlorophenol
Production by Decarboxylation of Natural Acids During Aqueous
Chlorination. Environ. Sci. Technol., 13, 325.

Lewis, I.C., Singer, L.S. (1981). Adv. Phys. Chem. Carbon, 17, 1.

Loper, J.C., Lang, D.R., Smith C.C. (1978). Mutagenicity of Complex
Mixtures from Drinking Water. Chapter in Water Chlorination:
Environmental Impact and Health Effects. Vol. 2, R.L. Jolley,
H. Gorchev, and D.H. Hamilton, Eds. Ann Arbor Sci. Pub., Ann
Arbor, MI., 433.

McCreary, J.J., Snoeyink, V.L. (1981). Reaction of Free Chlorine
with Humic Substances Before and After Adsorption on Activated
Carbon. Env. Sci. Technol., 15, 193.

McCreary, J.J., Snoeyink, V.L., Larson, R.A. (1982); A Comparison of
the Reaction of Aqueous Free Chlorine with Phenolic Acids in
Solution and Adsorbed on Granular Activated Carbon. Environ.
Sci. Technol., 16, 339.

N.A.S. (1979). An Evaluation of Activated Carbon for Drinking Water
Treatment. N.A.S. Subcommittee on Adsorption of the Safe Drinking
Water Committee. Board of Toxicology and Environmental Health
Hazards. Assembly of Life Sciences. N.R.S., Washington, D.C.

Norwood, D.L., Johnson, J.D., Christman, R.F., Hass, J.R., Bobenrieth,
M.J. (1980). Reactions of Chlorine with Selected Aromatic Models
of Aquatic Humic Material. Environ. Sci. Technol., 14, 187.

Pelon, W.B., Whitman, B.F., Beasley, T.W. (1977). Reversion of Histidine-
Dependent Mutant Strains of Salmonella Typhimurium by Mississippi
River Samples. Environ. Sci. Technol., 11, 619.

Puri, B.R., Singh, G., Sharma, L.R. (1957). Acidoid Behavior of
Charcoal as a Function of its Oxygen Complexes. Part 1. Electro-
metric and Conductometric Titrations With Alkalies. J. Ind. Chem.
Soc., 34, 357.




72









Puri, B.R., Singh, D.D., Chander, J., Sharma, L.R. (1958a). Interaction
of Charcoal with Chlorine Water. J. Ind. Chem. Soc., 35, 181.

Puri, B.R., Sharma, L.R., Singh, D.D. (1958b). Studies in Catalytic
Reactions of Charcoal, Part 1. Catalytic Oxidation and Decomposition
of Salt Solutions. J. Ind. Chem. Soc., 35, 765.

Registry of Mass Spectral Data. (1974). E. Strenhagan, S. Abrahamsson,
F.W. McLafferty, Eds. Vol. 1-4. John Wiley & Sons, New York, N.Y.

Rideal, E.K., Wright, M.W. (1926). CCXL. Low Temperature Oxidation at
Charcoal Surfaces. Part II. The Behaviour of Charcoal in the
Presence of Promoters. J. Chem Soc., 128, 1813.

Snoeyink, V.L., Lai, H.T., Johnson, J.H., Young, J.F. (1974). Active
Carbon: Dechlorination and the Adsorption of Organic Compounds.
Chapter in Chemistry of Water Supply, Treatment, and Distribution.,
A.J. Rubin, Ed. Ann Arbor Sci. Pub., Ann Arbor, MI, 233.

Standard Methods for the Examination of Water and Wastewater (1980).
15th edition. American Public Health Association, Washington, D.C.

Swinarski, A., Rozwadowski, M., Siedlewski, J. (1970). Effect of
Modification of the Active Carbon Surface by Ferric and Magnesium
Ions on Sorption and Catalytic Properties of the Carbons in
Reactions with Hydrogen Sulfide (Pol). Chem Stostow., Ser. A., 13,
413: Chem. Abs. 11510n, 1972.

Voudrias, E.A., Dielmann, L.M.J., Snoeyink, V.L., Larson, R.A., McCreary,
J.J., Chen, A.S.C. (1982). Reactions of Chlorite with Activated
Carbon and with Vanillic Acid and Indan Adsorbed on Activated
Carbon. Accepted for publication in Water Research.

Yohe, T.L., Suffet, I.H. (1978). Presented before the Symposium on
Activated Carbon Adsorption of Organics from the Aqueous Phase,
ACS, Div. of Env. Chem., Miami Beach, FL.
















APPENDIX



Possible Structures for the High Molecular Weight Compounds Present
in the Influent Solutions Collected on Carbon



Aqueous Vanillic Acid Chlorination Products


Compound #1 (Compound 12 in Table 5)

Molecular Weight: 338

Retention Time: 49.9 minutes

Electron Impact Spectrum:


I Iii! .1. ii I


14 0


I


1.0 I


I I


S240 2'


0 :340 39 .


440 4.1 "
440 4'Li 540


rn/el ~fTiU


.* rel
abun :


40 90


100_
. rel
S-buInd -


I 1 111


- II /il i.~_~


I











(OCH3) (CH)


Possible Structure:


Cl


The fragment at 171 could correspond to:


(m + n)=5


(OCH3)




Cl


Therefore m could be equal to 2 and n equal to 3 to give the general
structure.


~ +









Compound #2 (Compound 10 in Table 5)

Molecular Weight: 352

Retention time: 47.7

Electron Impact Spectrum:




%. re,
b u ri : 1

.5


Ii l1li,,l, lJ I I I iAl i U l I I: i i i 1-|! 1 1 1 I


I d11 Ii I I ,,,


100
. rel
ab u -i d

50


40 90 140 1.. 24,' ,I










I II
1


29 0 C 440 40 540
( ntr ."1 u




Possible Structure:


(OCH)m (OH3)n




Cl CH3
S(m + n)=5










Compound #3 (Compound 11 in Table 5)

Molecular Weight: 386

Retention time: 48.3

Electron Impact Spectrum:


Sre

5-


40 C0 140


188

rel
b lu n d






29S :348


Possible Structure:


1'y0 240 2' 0


* p ii117FF


38 440 4 I 5 40
tr ..' .. u


(OH) (OH)





Cl CH3 Cl
(m + n)=5


i /


' I I I 1 1 1 1 I '" "''~'~~'' ''" '- -~' ~' '~~


' ~' '~"'


I


t . L !. .


, i










Aqueous Ferulic Acid Chlorination Products


Compound #1 (Compound 17 in Table 9)

Molecular Weight: 380

Retention time: 45.9

Electron Impact Spectrum:


Sre 1
tab u r d-

560










*-0 rel
-
+:,I- u P+

E:B.


2:?0


Q ]L~I,


In' h H


i :: 1: I


L I


...I


I eI 3. L a
( rri .," @ .=m u


Possible Structure: No structure could be derived-from this spectrum.


1 2k


I


I I 0


S"" I L. I
43::: 0











Compound #2 (Compound 38 in Table 9)

Molecular Weight: 380

Retention time: 47.2 minutes

Electron Impact Spectrum:


r el


a i -












" rel
.~b und r

5S-


:1.,' ..


I 'I 1 0


-r '' -''' '' "" ~ -- --~ T T ~ ~ ~ ly I I ii1V


I i L I


2 :s: 0


S3: 3e0
( m|" e amru


Possible Structure: No structure could be derived from this spectrum.


I,


I""i~;'lLi


111


I


I I II i I II,


288


I


4 : 0 4 '
4:94:' 0O










Compound #3 (Compound 41 in Table 9)

Molecular Weight: 408

Retention time: 51.7 minutes

Electron Impact Spectrum:





100
1 rel
*ab 1 rd -l


100
re -
aburd' 1
6-
EO-~i


S[I11 1 1 Ill


230


If I I


I li l I I,


'2 : e '


I 3I 'i
( mi /.) .am


" "' "' "" i ~ ; "11 111 I II LI 1


4 C0
.... I ..L
4 ::


I i 3 I' 0L .
3 3 L


No structure could be derived from this spectrum.


, I,


Possible Structure:









Compound #4 (Compound 39 in Table 9)

Molecular Weight: 428

Retention time: 49.5 minutes

Electron Impact Spectrum:


180
re1






*-rb u r-~ ci
bun.



6 1,





r I,_


ll ,, ,,, .,l."_L_.. I I,


Sni'e)r aru






Possible Structure:

(OCH 3 (0 H3 n





Cl (m + n)=8











Compound #5 (Compound 40 in Table 9)

Molecular Weight: 428

Retention time: 51.4 minutes

Electron Impact Spectrum:


,, L i I. ,


I I


i i .L i i. .


4me) 4-'
m4.'e 1 1 r4u


Possible Structure:


(OCH ) (OCH3)n






Cl (m + n)=8


These high molecular weight compounds were collected on carbon after
the influent solution was reduced. Voudrias et al. (1982) reported a
series of hydroxylated biphenyls in the chlorination of p-chlorophenol
on carbon. It is apparent from the results in this report that chlorine
may not be necessary in the influent solution for reactions of this type
to occur.


abound

5r0.











10S
t.u re 1
abud ,-i


'-'-4


.1 .


ii


2 s


\ J 1 II l i


~


, i ,


~rr I II II Ilrrrl I r Imrllirrrl lr1717 II I .'1 _111


.5~la


i ~ ~ ~ ~ ~ i" --' "- '.i. .i .i. . .i '


'




Full Text

PAGE 25

N o 20 25 10 7 Figure 10. Total Jon Current Chromatogram of an F-400 (GAC) Extract of the Vani11ic Acid/Chlorine/Carbon Reaction.

PAGE 32

COOH COOH COOH. COOH COOH ___ o:::OH. 2 OH C1 COOH COOH COOH C1 OH (15) OH (27) C1 JD1CH3 .-"---OH OH OH (19,21,23, (14,16,lB,20 C1 24,25) C1 22) C1 OH (8) OH (4) C1 (OH)2'6 (OH)2 OH C1 I CH ... --3 OH (12) OH (7 13) OH (5) OH (2,6) .*rnethy1ated derivatives were actually found Figure 16. Reaction Scheme for the Vanillic Acid/ Chlorine/Carbon Reaction. o o OCH3 o 0 o OH y,. o (5) 0 (6) O' &,:(OHl 2 Cl, (OH)2 OCH 3 o (7,13) 0 (12) AjJOCH3 -OH OH (28)

PAGE 33

Compounds 1, 3 and 11 (Table 7) could not be identified. Compound 1 had a molecular weight of 146 and was not chlorinated. Compound 3 had a molecular weight of 174 with a fragmentation pattern of m/e 159, 143, 115, and 69. Its area was 3.5% of cumulative area of all the products formed. Finally, compound 11, a non-chlorinated aromatic compound, had a molecular weight of 204. The major products formed in the reaction are shown in Figure 15. Methyl, trimethoxy benzoate (20) and its monochloro derivative (2S) represented more than SO% of the area occupied by all compounds. Methylated vanillic acid (10) and its monochloro derivative (15) occupied an area of lS.4% and 3.9% respectively. Compound 7, a tetramethoxy benzene or trimethoxy quinone, had an area of S.2%. A general reaction scheme for the vanillic acid/chlorine/carbon reaction is shown in Figure 16. 3.3. Aqueous of Ferulic Acid The chlorination products of ferulic acid were determined using a blank reactor column. The experimental parameters, similar to those used in the chlorination of vanillic acid are presented in Table 8. The effluent of the c6lumn was reduced with sodium sulfite and pumped through a 3-inch bed of F-400 granular activated carbon. The carbon was then removed and Soxhlet extracted to give the total ion current chromatogram shown in Figure 17. This chromatogram is much more complex than the one shown in Figure 8 for the aqueous chlorination of vanillic acid under the same conditions. A list of the compounds tentatively identified in the aqueous extract and percent area in the chromatogram is presented in Table 9. It is obvious that the aqueous chlorination of ferulic acid yielded many more products than the vanillic acid chlorination. Phenol (3) and its monochloro, dichloro, and trichloro derivatives (S, 8, 9) were present in the extract in trace quantities. Phenol may have been present as a contaminant in either the water or the ferulic used in the experiment. Monochloro trimethoxy phenol (12) and dichloro dimethoxy benzene (10), chlorination products of vanillic acid, were also found. A series of carboxylic acids were extracted from the activated carbon. Methyl benzoate (4), methyl, dimethoxy benzoate (16), methyl, monochloro and dichloro dimetnoxy benzoate (18, 19) and methyl, tetramethoxy benzoate (24) were among the methylated carboxylic acids identified. Two isomers of methyl, monochloro methoxy benzoate (13, 14) were present at 2.8% and 2.7% of the total area in the chromatogram. Methyl, dichloro dimethoxy benzoate was present at an abundance of 2%. 28

PAGE 34

", Table 8. Experimental Parameters for the Feru1ic Acid/ Chlorine Blank Reactor Column Column diameter, cm Column length, cm Influent feru1ic acid conc., mg/1 Influent chlorine conc., mg/1 Influent pH F1owrate,-1/m2-min Reaction time, min Reaction volume, m1 Total volume passed through bed, 1 29 1.9 35. 2.6 10.0 6.0 69.1 2.0 38.6 18.8

PAGE 35

w o 12 I i 6 I 2 51 1q1111 1 3 I 8 Figure 17. Total Ion Current Chromafogram of an F-400 (GAC) Extract of the Aqueous Chlorination of Ferulic Acid

PAGE 43

, ** SFECTRUM DISPLAY/EDit FRII 9131 .. sPECfilul'l bTsptR'l7tD IT .. rRH 9131 RUU C UItVCL E:lAHK fWIt F'HEi.;,oZUL+ J S 1ST SC/PGI IRIII: RUII PH**2UL+lS 1ST SC/PGI 414 X.50 y. X.50 y. I.OO"KRB"30-280@5/MIH 2 1 3 5 1 6 I 7 4 III 1 ) _--vJ ....J IJ '-v. 11 "I ,,> 2'3 2'4 2'5 2'" 2'7 2'e 2'9 ,'. 3 32 33 3'4 ,\< ,'. ,,, 41. .', 4'> E 44 4'< 47-49 5'9 5', 5'2 Figure 19. Total Ion Chromatogram for the Methylene Chloride Extract of the Aqueous Chlorination of Ferulic Acid.. 1 s;t '. 849 X-.50 y. Lee ..

PAGE 44

CR= CRCl CR= CRCl ( 1) tCl (5) OCR3 OCR3 I OCR3 OCR3 0 II (2 ) @-
PAGE 45

Although the chlorinated styrenes were found in large quantities in the aqueous extract of chlorinated ferulic acid, these products were not found in the aqueous extract collected on carbon. These compounds were apparently hydroxylated via a free radical mechanism at the carbon surface. Trimethoxy and tetramethoxy B-chlorostyrenes were recovered off the carbon (Table 9). It is significant that free chlorine residual was not required for this reaction to occur. It was determined from these two experiments that a liquid-liquid extraction technique provided more accurate results for the reaction products of ferulic acid and aqueous chlorine in the absen.ce of carbon. The most common reactions in the aqueous chlorination of ferulic acid were oxidative decarboxylation of the alkyl side chain with concomitant substitution of chlorine, oxidation of the benzylic carbon to form acetophenones, and substitution of chlorine into the aromatic ring. 3.4. Ferulic Acid/Chlorine/Carbon Reaction The catalytic effect of activated carbon on the reaction of ferulic acid and chlorine was studied using F-400 GAC. The experimental parameters are listed in Table 11. A total ion current chromatogram (electron-impact mode) of the methylated extract of the activated carbon used in the experiment is shown in Figure 21. The extract was also run in the Negative Ion -Chemical Ionization mode (NI/CI) as an aid in determining which compounds were chlorinated. Negative Ion -Chemical Ionization also enabled the molecular weight to be determined for those chlorinated compounds which had a weak molecular ion in the electron impact mode. The NI/CI chromatogram is shown in Figure 22. Table 12 presents the tentatively identified compounds in the extract of the ferulic acid/chlorine/carbon reaction. The compounds are listed in the order of elution in the total ion current chromatogram (Figure 21) with percent areas in the chromatogram. Most of the compounds tentatively identified in this reaction were present in the influent to the carbon column. A majorseries of products; monochloro compounds of molecular weight 204 and the dichloro derivatives of molecular weight 238 were intriguing. A logical structure could not be derived from the fragmentation pattern of the low resolution mass spectra .. The electron impact spectra demonstrated a loss of -Cl or -OCH3 to yield a large M-35 and M-31 fragment ion. The molecular ion could be confirmed only through the use of NI/CI spectra (Figure 23). The spectra of the MW 238 compound was similar, with a large M-35 ion and another fragment corresponding to a loss of -OCH3. Although these compounds comprised approximately 30% of the total area, they were not present in the influent to the column and were not previously reported chlorination products of 40

PAGE 46

Table 11. Experimental Parameters for the Feru1ic Acid/Chlorine/Carbon Reaction Column diameter, cm Column length, cm Influent Feru1ic Acid Conc., mgl1 Influent Chlorine Conc., mgl1 Influent pH Flowrate, 11m2-min Reaction time, min Reaction volume, m1 Total volume passed through bed, 1 Type of water used 41 1.9 38 2.6 10.0 6.0 77 .6 2.1 45.8 132. di sti 11 ed

PAGE 47

. N Figure 21. Total Ion Current Chromatogram of an F-400 (GAC) Extract of Ferulic Acid/Chlorine/Carbon Reaction

PAGE 48

.po w I Figure 22. Negative Ion Chemical Ionization Chromatogram of an F-400 (GAC) Extract of the Ferulic Acid/Chlorine/Carbon

PAGE 57

U1 N 1 2 1. S. 3 4 Figure 27. Total Ion Chromatogram for the Ferulic Acid/Chlorine/Carbon Reaction (Second Experiment).

PAGE 58

o / "'>-0 CH3 0 C -CH3 o II CR= CR-CR (01 OCR3 OCR3 o II C-H OCR3 o II C1 tC -CR3 .OCR3 o 1\ C-CR I 3 Table 17. Compounds Tentatively Identified in the Ferulic Acid! Chlorine/Carbon Reaction (Second Experiment). 53

PAGE 59

3.5 Aqueous Chlorination of Ferulic Acid at pH = 10. In order to study the aqueous reaction products of ferulic acid and chlorine under normal water treatment plant conditions, ferulic acid was chlorinated at a pH of 10 and the reaction products were examined. At this pH, the predominant chlorine species is whereas the previous work had HOCl as the predominant chlorine species. Chlorination of ferulic acid at high pH might occur in a water treatment plant during the softening process and evaluation of the effect of pH on this reaction would allow determination of the optimal point of chlorination. Hypochlorous acid is an electrophilic reagent at either the chlorine or the oxygen atom. At a pH greater than approximately 8.0 (Snoeyink and Jenkins, 1980), chlorine residual exists primarily as hypochlorite anion (OCl-). It has generally been observed that more rapid addition of chlorine occurs at pH regions where HOCl predominates. Compounds formed through chlorination followed by base-catalyzed hydrolysis may occur in higher yield at higher pH, however (Boyce et 1983). The chlorination products of ferulic acid at a pH of 10 were determined using a blank reactor as in previous experiments. The experimental parameters are shown in Table 17. Both influent solutions to the column were adjusted to pH = 10 with NaOH pellets and buffered with 0.001 M phosphate. The effluent from the column was collected in a 40 liter carboy and a liquid-liquid extraction was performed on a 5 liter sample. The total ion current chromatogram is presented in Figure 28. Tentatively identified compounds and their percent areas in the chromatogram are shown in Table 18. Both ferulic acid and chloro ferulic acid were found as products in this reaction. The presence of these compounds was expected and agreed with findings by Norwood (1980) for chlorination at lower pH. Several isomers of dimethoxy B-chlorostyrene and chlo.ro dimethoxy B-chlorostyrene were found in high yield. These two compounds were a 1 so found by Norwood et Acetophenones were another group of products formed in the aqueous chlorination of ferulic acid at pH 10. Chloro methoxy acetophenone and chloro dimethoxy acetophenone were both found in moderate yield in this experiment. These compounds were found in all previous experiments. Thus, pH appeared to exert a minimal influence on the formation of these compounds. 54

PAGE 60

Table 17. Experimental Parameters for the Aqueous Chlorination of Ferulic Acid at a pH of 10.0. Column Diameter, cm Column Length, cm Influent Ferulic Acid Cone., mg/l Influent Chlorine Cone., mg/l Influent pH Flowrate, ml/min Reaction time, min Reaction volume, ml Total Volume Passed, 1 Type of Water Used 55 0.9 29.7 3.0 10.55 9.95 11.19 1.69 18.9 5. tap

PAGE 61

, I 4 l( ---.l.LZ! i? 23 24 1', f'6 4 2 FRN tio 1ST SC'PG. 4 >c-.S0 V-1. 1 3 Figure 28. Total Ion Chromatogram for the Aqueous Chlorination of Ferulic Acid at pH = 10.0

PAGE 62

0 II C-CH CH=CHCl I 3 Cl 8 1,"2 Cl OCH3 OCH3 OCH3 0 0 1/ C-CH II I 3 CH = CH C OCH3 Cl 9,10 3 ./ 3 OCH3 0 CH=CHCl II C-H 11 4,5,6 OCH3 OCH3 OCH3 0 0 \I C-CH3 7 OCH3 OCH3 OCH3 Table 18. Tentatively Identified Compounds From the Ferulic Acid! Chlorine/Carbon Reaction at pH = 10.0 57

PAGE 63

3.6 Ferulic Acid/Chlorine/Carbon Reaction at pH 10 The effect of pH on the ferulic acid/chlorine/carbon reaction was examined by adjusting the pH of influent solutions to 10 and buffering these solutions with 0.001 M phosphate. The experimental parameters are presented in Table 19. The total ion current chromatogram of the methylated extract of the activated carbon used in the experiment is shown in Figure 29. Table 20 presents tentatively identified compounds in the extract of the ferulic acid/chlorine/carbon reaction. Products are listed according to abundance. 3.7 Aqueous Chlorination of a Natural Water Sample Previous experiments were conducted for the purpose of understanding the types of reactions which occur during aqueous chlorination of compounds commonly found in natural waters. Compounds were selected to determine the types of reactions which occur at specific functional groups in order to provide a basis for the analysis of chlorination products of naturally occurring humic and fulvic acids. These compounds are of interest due to their ubiquitous nature in and ground waters. Humic and fulvic acids are too complex to characterize individually and many researchers have used model precursor compounds to investigate specific reaction chemistry. Carbon-catalyzed chlorination products are especially important with humic and fulvic compounds as these products can be expected to occur in water treatment plant processes. A natural surface water was obtained which had passed through a cypress dome wetland and had been characterized by previous researchers (Dierberg et al., 1980). The particular water chosen was Hatchet Creek in Florida. This water was diluted with distilled make-up water to achieve a humic acid concentration of approximately 5 mg/l derived from total organic carbon (TOC). The TOC of humic acid was based on a correlation between color and TOC developed for this water by Dierberg et al.(1980). The influent solutions were buffered with 0.001 M phosphate and reacted in a blank reactor column as in previous experiments. The experimental parameters are listed in Tables 21 and 22. The total ion current chromatogram shown in Figure 30 indicates the presence of relatively few products. Among the chlorinated products found were chloroform, dibromochloromethane, and bromoform. The total ion chromatogram for the humic acid/ chloY'"{ne-/ ca rbon reaction is shown in Figure 31. In this chromatogram, several compounds are present which were not found in the aqueous chlorination of this natural water. Several compounds identified were phthalates and fatty acid methyl esters. A mass reconstruction was performed on masses 74 and 149 and is shown in Figure 32. The mass 149 is indicative of phthalates and the mass 74 is indicative of acid esters. Due to the 58

PAGE 64

Table 19. Experimental Parameters for the Reaction of Ferulic Acid and Chlorine on Carbon at pH = 10.0 Column Diameter, em Column Length,cm Influent Ferulic Acid Cone., mg/l Influent Chlorine Cone., mg/l Influent pH Flowrate, ml/min Reaction time, min Reaction volume, ml Total volume passed through bed, 1 Type of water used 59 0.9 29.7 3.0 10.55 9.95 5.80 1.61 9.39 60. tap

PAGE 65

C'\ o ** SF'[ .lJI1 [I ,.f'L 'r,' II ** fill ':1 ** 9 9 .it SF' UI1 I S LHY" DI ** fRtI l1 C I t!tvCL EXTRACT -PHz: 1 O+HiUL LE W J TH IS: (HWL IIGTUwC/f'G I IRBOU EXTRACT-PHa 10 .... SAHPlE WI TH IS (1 eUL I 41lR80H EXTP.ACT-PH= 1 a* .... 2UL S',AI'1f'L( WITH IS' lOUL IISHIGC/P(; I S23 IO.'22/SZ ... k.:F:B**30-2O@5d1JtI WITH 20 11IH X" .50 y. i.0KR:8**30-ZSJ@SdllN WITH.c:0 tUN X-.'S0 y. 1.0IKRB**30-280QlS/MUI t.IITH 20 I'1lN Xz y1 Ti '4 'I 34 Figure 29. Total Ion Chromatogram for the Ferulic Acid/Chlorine/Carbon Reaction at pH = 10.

PAGE 70

en 01 S EeT U D LA/E HHTCHET CREEtUCL EXT"ZUL+IS KRE**3/1,/e3 30-2S0G5/MIH TJ .. fRH 5 SPE I L I 1ST SC/PG, I CARBON EXT"ZUL+IS Xc' .50 y. 1.09 l/83**30-2S0aS/I1IH .. fRH 95 6 ** SPECT UI1 D L I 1ST SC/PG, 433CREEK/CL CRRBOH EXT"2UL+IS Xc .60 V1.00 31. Total Ion Chromatogram for the Humic Acid/Chlorine/Carbon Reaction ** RN 9 2 1 Sf SC/PG, 866 X-.60 Y-1.00 .-

PAGE 73

NEC' TF-un fl /( II I T ** SPEer RUM lH sF-LAY /EIt [TT ".:;;.----r"'F"':u;--9"'5"2"'6-------..:;;*'SPECTRUt" D I /ED I T .. (PEEl, CL E::T..,ior2UL-tt: SC/F'I;, I CF:EEI( .... C'L (AF'[:OU 15.T :SC/F'Gg .,4El CREEIU'("L CAF'EOII [in1t-*2UltIS Pl: 3' 1 :;.:z:, ..... tl :.;,. so )':::1 1 .00, '51 X=- '(a 1 19/e3"'*3-0-Z.E:1j1j5JIHU nil I Sf $C/PG I aS3 j{. .50 Y= 1.(ji) g: 1 I __ __ _l. \. ,1'e," J,{II .. '. ,I,"\, ,,",SI ".\ ,b I, ",; -=,'1::'7-:131.:---::3:1:'<;-":'1.',':", -.j:5-4T'.--'.\;:-, -.T.',;-"".C;C:-, Figure 32. Mass Reconstructed Chromatograms for the Humic Acid/Chlorine/Carbon Reaction