UV direct photolysis of N-nitrosodimethylamine (NDMA): Kinetic and product study

Article abstract of DOI:10.1002/1522-2675(200205)85:5<1416::AID-HLCA1416>3.0.CO;2-I

The ultraviolet (UV) direct photolysis of N-nitrosodimethylamine (NDMA) in aqueous solutions at pH 3 and 7 leads to dimethylamine, and nitrite and nitrate ions as the major degradation products. In addition, small amounts of formaldehyde, formic acid, and methylamine are formed. When the initial concentration of NDMA was 1 mM, only a 13% decrease in the total organic carbon (TOC) was measured at pH 7, whereas no significant change in the TOC was observed at pH 3. In the concentration range 0.01 - 1 mM NDMA, zero-order kinetics is obeyed, whereas first-order kinetics is followed at concentrations below 0.01 mM. The photolysis occurs much faster at pH 3 than at pH 7, which is explained by the difference in the quantum yields of the process at these two pH values. UV Direct photolysis is an efficient process for the removal of NDMA from contaminated waters, and electrical energy per order (E_EO) values as low as 0.3 - 0.5 kWh/order/m³ were calculated for treatment of low concentrations of NDMA (0.001 mM).

Full text of DOI:10.1002/1522-2675(200205)85:5<1416::AID-HLCA1416>3.0.CO;2-I

1
416
Helvetica Chimica Acta ± Vol. 85 (2002)
UV Direct Photolysis of N-Nitrosodimethylamine (NDMA): Kinetic and
Product Study
1
by Mihaela I. Stefan* ) and James R. Bolton
Bolton Photosciences Inc., 628 Cheriton Cres., NW, Edmonton, AB, Canada T6R 2M5
Dedicated to Professor Andr e¬ M. Braun on the occasion of his 60th birthday
The ultraviolet (UV) direct photolysis of N-nitrosodimethylamine (NDMA) in aqueous solutions at pH 3
and 7 leads to dimethylamine, and nitrite and nitrate ions as the major degradation products. In addition, small
amounts of formaldehyde, formic acid, and methylamine are formed. When the initial concentration of NDMA
was 1 mm, only a 13% decrease in the total organic carbon (TOC) was measured at pH 7, whereas no significant
change in the TOC was observed at pH 3. In the concentration range 0.01 ± 1 mm NDMA, zero-order kinetics is
obeyed, whereas first-order kinetics is followed at concentrations below 0.01 mm. The photolysis occurs much
faster at pH 3 than at pH 7, which is explained by the difference in the quantum yields of the process at these two
pH values. UV Direct photolysis is an efficient process for the removal of NDMA from contaminated waters,
3
and electrical energy per order (EEO) values as low as 0.3 ± 0.5 kWh/order/m were calculated for treatment of
low concentrations of NDMA (0.001 mm).
Introduction. ± N-Nitrosodimethylamine (NDMA) is not an industrially or
commercially important chemical, but its occurrence in the human environment, as
detected in outdoor air, surface waters, and soil, comes from its potential release from
industries, such as tanneries, amine, pesticide, and detergent manufacturing plants,
rubber and tire manufacturers, fish and food processing industries, malt beverage
industries, dye manufacturers and cosmetics industries. It is also a hydrazine-
decomposition intermediate. In water and soil, NDMA is formed by the reaction of
widely occurring primary, secondary, or tertiary amines with nitrites. It has been found
in groundwaters at levels as high as 10 ppb, and there is a growing concern with regard
to health effects associated with exposure to NDMA.
Hepatotoxicity is the primary systemic effect of NDMA, and the mechanism of
NDMA-induced liver damage may be related to the alkylation of cellular proteins [1].
Genotoxicity of NDMA in vitro and carcinogenity in various species of animals have
been demonstrated in numerous studies over the past two decades.
Recent literature reports emphasize that both artificially generated UV light or
natural sunlight are sufficient to enable NDMA to induce mutagenicity by methylation
of DNA [2][3]. Consequently, NDMA has been assessed as a potential carcinogen in
humans.
N-Nitrosodimethylamine does not currently fall under the U.S. EPA regulations,
but it is listed as a priority pollutant on the U.S. EPA National Priorities List, and
À5
regulations and advisories are enforced in various states. At a 10 cancer risk level set
1)
Current address: Trojan Technologies Inc., 3020 Gore Road, London, Ontario, Canada N5V4T7.

Helvetica Chimica Acta ± Vol. 85 (2002)
1417
by the U.S. EPA among the ambient-water-quality criteria, the corresponding level for
NDMA would be 0.0014 mg/l, and the EPA drinking water standard at the same risk
level would be 0.0007 mg/l [4]. The Ontario Ministry ofEnvironment , Canada, established
an interim drinking water standard for NDMA of 0.009 mg/l (9 ppt). The California
Department ofHealth Services has set an interim control level of 20 ng/l (20 ppt).
It is known that N-nitrosamines are thermally stable in aqueous solutions. For
example, it was found that their hydrolysis in aqueous HCl after refluxing for 24 h did
not proceed beyond 10 ± 12% [5]. Also, they are fairly resistant toward biodegradation,
do not air strip or adsorb on activated carbon. However, they are photochemically
labile, and, therefore, advanced oxidation technologies, based on irradiation with UV
light, have been promoted for the removal of NDMA from contaminated waters.
The first study the NDMA photolysis was published in 1939, when Bamford [6]
decomposed NDMA in the gas phase with both medium- and low-pressure Hg lamps.
Several gaseous products, such as H , CH , N , CH ˆCH , and Me NH (DMA) were
2
4
2
2
2
2
identified. Since then, several other studies on the photolysis of NDMA in the gas
phase have been reported in the literature, and mechanistic pathways were attempted
[
7 ±9].
The acid-catalyzed photochemistry of nitrosamines in aqueous solution, including
NDMA, has been thoroughly investigated by Chow and co-workers [10 ± 13], whereas
fewer studies on this subject have been conducted in neutral media [14] or in strong
alkaline solutions [15]. Ho et al. [16] found a relatively strong pH dependence of the
quantum yield for the direct photolysis of NDMA.
Mechanistic studies on the photolysis of NDMA indicated that the principal
À
products are DMA and NO2 ions. Small amounts of N O, N , HCHO, methylidene-
2
2
methylamine, amidoxime derivative, MeNH , and HCOOH can be also formed, but
2
their levels are strongly dependent on the environment [11 ± 14][17 ± 24]. Adeleke and
Wan found ESR evidence for nitric oxide as a primary product in the photolysis of
NDMA in aqueous solution at room temperature [25], whereas Jakubowski and Wan
identified the dimethylaminium radical as an intermediate radical species, by ESR
spectra at 77 K [26].
Over the past decade, UV companies have installed several full-scale UV reactors
in North America for treatment of NDMA in drinking water and in ground and
wastewaters. The concentration of NDMA in the influents can be as high as 100 ppt in
drinking water, 1.5 ppm in groundwater and 1,400 ppm in wastewater. Ozone alone or
ozone/UV processes are totally inefficient in the degradation of NDMA [27]. Similarly,
preliminary studies on the UV/H O process applied to the NDMA removal indicated
2
2
no significant improvement of that treatment compared to UV alone [28][29].
Even though some mechanistic pathways have been proposed in the literature for
the direct photolysis of NDMA [12 ± 14][30], to our knowledge, no thorough product
or kinetic study, and/or organic carbon and nitrogen balances are reported in the
literature.
Here, we report results on the kinetics and degradation products of the UV direct
photolysis of NDMA at pH 3 and 7 in dilute aqueous solutions, and discuss briefly a
reaction mechanism based on the identified and quantified intermediate by-products.
The degradation profiles of NDMA and its intermediates are correlated with the
organic carbon and nitrogen mass balances.

1
418
Helvetica Chimica Acta ± Vol. 85 (2002)
Results and Discussion. ± The NDMA absorption spectrum exhibits a strong band
À1
À1
with a maximum at l ˆ 228 nm (e ˆ 7378 m cm [31]), and a weak band with a smooth
À1
À1
maximum at l ˆ 332 nm (e ˆ 109 m cm [31]). The first absorption band is regarded
as a p ! p* intramolecular charge transfer, whereas the second is associated with an
n ! p* transition band. When irradiated with UV light, the NÀN bond is broken in the
NDMA molecule, generating reactive radicals that are involved in further degradation
À
À
processes. Fig. 1 shows the absorption spectra of NDMA and NO /NO ions. The
2
3
latter are among the major degradation by-products of NDMA, which, once released
into the irradiated system, will compete for the light with the parent pollutant.
Fig. 1. Absorption spectra ofNDMA and its major degradation products, and the spectral distribution ofthe
1-kW medium-pressure Hg lamp
Degradation Products during the UV Photolysis ofNDMA . The irradiation of a ca.
1
mm aqueous NDMA solution at pH 3 led to a rapid decay of NDMA and the
À
formation of DMA and NO2 ions as the major products. HCHO was a minor
degradation product, which eventually was oxidized to HCOOH. NO Ions were
generated along with the NO2 ions, but at a lower level. The decrease of the NO2
concentration after the removal of NDMA is associated with the increase of NO3
concentration through a photo-oxidation process. CH NH is an expected degradation
À
3
À
À
À
3
2
product of NDMA, as being formed in a 1:1 ratio with HCHO. However, its
quantitative determination has not been attempted.
Fig. 2 shows the time profiles of NDMA and its degradation by-products during the
photolysis at pH 3. One observes that DMA reaches a maximum concentration of ca.
0.9 ± 1 mm at the same time when NDMA was fully photodegraded, and it does not
undergo any significant decay under further irradiation, since it does not absorb UV
light. From the industrial point of view, this might be a problem, when UV direct

Helvetica Chimica Acta ± Vol. 85 (2002)
1419
Fig. 2. Time profiles of NDMA and its degradation products, and TOC mass balance during irradiation at pH 3.
: NDMA; ~: dimethylamine;
: nitrite; ~: nitrate; Â : formaldehyde; : formica ci d; *: calculated TOC/2; *:
measured TOC/2. The −calculated× TOC represents the TOC in NDMA and degradation products.
&
&
*
photolysis is used to treat NDMA, since the persistence of DMA in the treated waters
can possibly lead to the re-formation of NDMA [32].
As shown in Fig. 2, there is a good agreement between the calculated and
experimentally measured total organic ca rbon (TOC) over the entire irradiation period,
which suggests that the major organic compounds present in the irradiated system were
properly identified and quantified. No significant decrease in the TOC level was
observed, which is consistent with the resistance of DMA to the UV treatment.
The degradation of 1 mm aqueous NDMA solution at pH 7 proceeded much slower
than at pH 3, which is expected given the difference between quantum yields
À
determined at these pH values [16]. DMA and NO ions were quantified as the major
2
products; HCHO was formed at a higher yield than in the case of photolysis conducted
À
at pH 3. However, at pH 7, NO ions were exclusively generated by the photo-
3
À
oxidation of NO2 ions, whereas at pH 3, as pointed out above, initially, they were
À
formed simultaneously with NO2 ions as primary degradation products. That may
suggest that the mechanistic pathways of the irradiated system are pH dependent. Fig. 3
shows the time profiles of NDMA, degradation by-products, and the TOC.
Similar to the experimental data obtained at pH 3, a good TOC balance at pH 7 was
found over the irradiation period, but a slight decrease of ca. 13% in the organiccarbon
was observed. CH NH might be also an intermediate, but it was not quantified.
3
2
No depletion of the oxygen concentration in the solution was observed during the
irradiation at either pH 3 or 7, which is consistent with the very low yield of
mineralization.

1
420
Helvetica Chimica Acta ± Vol. 85 (2002)
Fig. 3. Time profiles of NDMA and its degradation products, and TOC mass balance during irradiation at pH 7.
: NDMA; ~: dimethylamine;
&
&
: nitrite; ~: nitrate; Â : formaldehyde; : formica ci d; *: calculated TOC/2; *:
*
measured TOC/2. The −calculated× TOC represents the TOC in NDMA and degradation products.
A comparison between the total nitrogen concentration as calculated from the sum
of the nitrogen content in the detected and quantified compounds in the irradiated
system, with the initial concentration of nitrogen, as present in NDMA, reveals no
change at pH 3, and a ca. 20% loss of nitrogen at pH 7. This is an additional indication
that different reaction pathways are probably followed by the system, depending on
pH. The loss of nitrogen at pH 7 might be accounted for mechanistically by the
formation and stripping of nitrogen gases, such as N O, NO, and N , since no significant
2
2
concentrations of nitrogen-containing organic compounds are missing from the
analysis, as indicated by the TOC balance.
Degradation Pathways in the UV Direct Photolysis ofNDMA . The broad-band
irradiation from the medium-pressure Hg lamp makes possible both types of electronic
excitation in the NDMA molecule, i.e., p ! p* (l ˆ 228 nm) and n ! p* (l
ˆ
max
max
3
32 nm). The primary step consists in the photocleavage of the NÀN bond (170 kJ/
mol), with the generation of the corresponding dimethylaminium radical and NO, in a
solvent cage× [33]. The generated radicals can recombine in the cage, reforming the
−
original NDMA molecule, or can escape from the cage, leading to various products.
Polo and Chow [11] proposed a simplified mechanism for the photodegradation of
NDMA in weakly acidic aqueous solutions, as shown in Scheme 1.
Two distinct pathways are proposed for the decay of the excited state of NDMA.
One implies the hydrolysis of the excited state, leading to the protonated form of DMA
(
7; pK ˆ 10.73) and HNO (8). The second assumes the rearrangement of the excited
a
2
state of NDMA to the N-nitrosodimethylammonium ion (2), followed by dissociation

Helvetica Chimica Acta ± Vol. 85 (2002)
1421
Scheme 1. Primary Processes in the NDMA Photodegradation in Aqueous Solution
to the dimethylaminium radical (3) and NO radical (4). HNO (5) and N-
methylidenemethylamine (6) are generated as a result of the reaction of radicals 3
and 4 in the solvent cage [33][34]. The only stable species in Scheme 1 are DMA and
À
HNO /NO ions, which were identified and quantified in this study as the major
2
2
degradation products of NDMA, both at pH 3 and 7.
Further reactions of the above species can be postulated to explain the photo-
degradation products of NDMA observed in this work. These reactions are given in
Scheme 2.
Scheme 2. Generation ofthe Intermediate By-Products during the Photolysis ofNDMA in Aqueous Solution

1
422
Helvetica Chimica Acta ± Vol. 85 (2002)
The acid-catalyzed hydrolysis of the immonium compounds to the corresponding
carbonyl and amine compounds is known in the literature [13][14]. Therefore, this can
be a route toward the formation of HCHO and CH NH in the irradiated system in a
3
2
1
:1 ratio. HCOOH can be generated through the slow oxidation of HCHO. Similarly,
À
À
NO ions are slowly oxidized to NO ions; once formed in the system, these ions will
2
3
compete for the light with NDMA, reducing the effectiveness of the treatment,
especially when starting with a high concentration of parent pollutant.
The dimethylaminyl radical is almost unreactive toward oxygen [7][17], but it
readily reacts with H-atom-donating compounds by H-abstraction, forming the
corresponding amine. It has been proposed that the dimethylaminyl radical may
abstract a H-atom from another NDMA molecule, leading to DMA and the C-centered
radical 7, which rearranges to methylidenemethylamine 6 by losing the NO fragment
[
21][22].
No attempt was made to measure the nitrogen-containing gases, but their
generation may explain the decay of the organic nitrogen content during the photolysis
at pH 7.
Kinetics ofthe UV Direct Photolysis ofNDMA . The decay of NDMA in aqueous
solutions of pH 3 and 7, under irradiation with a medium-pressure Hg lamp, was
investigated over a wide range of concentration (0.001 ± 1 mm). Depending on the
initial concentration, the NDMA decay obeyed either zero- or first-order kinetics. In all
cases, the degradation of NDMA was followed for more than one order of magnitude.
Figs. 4 and 5 illustrate the kineticbehavior of NDMA solutions, at pH 3 and 7,
respectively.
Fig. 4. Kinetics ofthe NDMA photolysis at pH 3
One observes that for similar initial concentrations, the rate of degradation of
NDMA is significantly faster at pH 3 than at pH 7. This can be explained by the
difference between the quantum yields at these pH values, namely 0.25 and 0.13,

Helvetica Chimica Acta ± Vol. 85 (2002)
1423
Fig. 5. Kinetics ofthe NDMA photolysis at pH 7
respectively [16], and possible NDMA-reformation mechanisms in neutral solutions.
The Table summarizes the kinetic data and gives the efficiency of the UV process, as
expressed by the figure-of-merit −Electrical Energy per Order× (EEO), introduced by
3
Bolton et al. [35]. The Electrical Energy per Order (kWh/order/m ) is defined as the
electrical energy [kWh] required to bring about the degradation of a contaminant by
3
one order of magnitude in 1 m of contaminated water or air.
Table. Summary ofthe Kinetic Data fr om the Photolysis ofNDMA
NDMA/mma)
Time/minb)
k
À1
k
1
/min
À1
E
EO/kWh/order/m3
% Fabsc)
0
/mm min
pH 3
1
1 ± 5
± 11
1 ± 2
0.164
±
0.074
±
1.31
±
±
1.04
±
90
70
7
0
.1
2
1 ± 2
1 ± 2
± 4
±
±
±
2.6
4
4
0.53
0.34
0.34
0
0
.01
.001
30
4
pH 7
1
0
10 ± 60
0.017
±
±
0.008
±
±
±
±
±
4.3
29.6
±
1.63
1.29
0.53
90
70
6
7
0 ± 70
0 ± 90
1 ± 12
4 ± 20
1 ± 4
0.32
0.05
±
0.84
1.063
2.6
.1
1
0
0
.01
.001
30
4
1 ± 2
a) Initial concentration of NDMA. b) Time interval during the irradiation of NDMA, when a specific kinetics is
c
obeyed. ) Fraction of light absorbed in the UV reactor (pathlength 8 cm).

1
424
Helvetica Chimica Acta ± Vol. 85 (2002)
At large initial concentrations (1 and 0.1 mm), NDMA degrades according to zero-
order kinetics, followed by first-order kinetic behavior after a significant depletion in
À
À
concentration. The formation of NO2 and NO3 ions, to a large extent, leads to a
decrease of the NDMA photolysis rate given the increased competition for light (see
spectra in Fig. 1). The two first-order decay regimes at pH 7 and the significant
differences between the first-order rate constants, characterizing the decay of NDMA
in the same ranges of concentration, may also be explained by an inner-filter effect of
À
À
the NO and NO ions. From the known spectral emission of the UV lamp, one can
2
3
calculate that 38.2% of the total UV output of the lamp in the 200 ± 300 nm range is
À
À
emitted below 250 nm, where both NDMA and the NO2 and NO3 ions absorb
significantly. Therefore, when a significant fraction of NDMA is depleted, the built-up
products can compete for the light, decreasing the rate of further degradation of
NDMA. For example, when starting with a 1 mm NDMA solution, pH 7, after 66 min
À
À
irradiation, the NDMA, NO2 and NO3 ion concentrations were 0.011, 0.71, and
0
3
.0436 mm, respectively. The corresponding fractions of light absorbed within 200 ±
00 nm are 12.5, 25.5, and 3%. When starting the irradiation with a 0.01 mm NDMA,
pH 7, the fraction of light absorbed by NDMA at t ˆ 0 min is 30%.
Similar correlations can be made with respect to the E values. Considering the
EO
low levels of NDMA encountered in either ground or drinking waters (ppb ± ppt), it
appears that UV direct photolysis is a very good treatment option.
Experimental Part
Supplies and Reagents. N-Nitrosodimethylamine (NDMA), anal.-reagent grade, was purchased from
Aldrich. All chemicals used for the standard solns. in the instrumental analysis were anal.-reagent grade and
used as received from various chemical suppliers.
UV Reactor. The UV irradiations were carried out in a semi-batch 1-kW UV Rayox ¾ reactor provided by
Calgon Carbon Corporation, Pittsburgh, VA, described in detail in [36]. A 1-kW medium-pressure Hg lamp was
used in all experiments. The total incident photon flow within 200 ± 300 nm wavelength range was (3.03 Æ
À4
À1
0
.08) Â 10 einstein s , as determined by persulfate actinometry [37]. The 1-kW medium pressure Hg lamp
requires a 50-s warm-up period to reach full power.
All experiments were conducted in aerated aq. solns. (distilled water); the pH was adjusted with NaHCO
3
or H
2 4
SO to the desired value.
Analytical Determinations. N-Nitrosodimethylamine (NDMA) Analysis. NDMA was measured by either a
GC or HPLC, depending on the investigated concentration range.
A GC method was used for all the experiments starting at an initial concentration of NDMA of ca. 1 mm. A
HP 6890 gas chromatograph equipped with an autosampler was employed with a HP Wax (cross-linked
polyethylene glycol) column (30 m  0.53 mm, 1-mm film thickness) with a temp. program, as follows: 358 for
5
min, then raised to 1508 at 258/min and held for 2 min at 1508. The injector temp. was 2108, with a split injection
ratio of 10 :1; the detector temp. (FID) was 2508. He was used as a carrier gas at a flow rate of 16.1 ml/min. The
corresponding retention time of NDMA was 8.61 min.
Concentrations as low as 0.005 mm NDMA could be accurately measured under the GC conditions stated
above, by means of a multiple point calibration curve.
An HPLC method was used for the low range of the NDMA concentrations. A Waters 510 HPLC system
with a Waters 486 tunable UV detector was used in connection with a Supelcosil LC-18 column (25 cm  4.6 cm,
2
Supelco). The mobile phase was MeCN/H O 60 :40 (v/v) for the experiments at pH 3, and 45 :55 (v/v) for those
3 4
at pH 7, acidified at pH 3 with H PO ; at the flow rate of 1.2 ml/min, the retention time of NDMA was 2.7 min.
The detection wavelength was set at 240 nm for most of the measurements, but shifted to 228 nm when
concentrations in mm and sub-mm range were analyzed.
Concentrations of NDMA as low as 0.015 mm could be measured accurately by the HPLC method described
above, based on a multiple point calibration curve.

Helvetica Chimica Acta ± Vol. 85 (2002)
Dimethylamine (DMA) Analysis. The reaction of DMA with CS
1425
2
in the presence of CuCl
2
leads
quantitatively to the yellowed-colored cupric salt of dithiocarbamic acid, which can be measured spectrophoto-
metrically at l ˆ 434 nm [38]. Under the reaction conditions, the primary and tertiary amines do not interfere.
This analysis is known as the Umbreit method. An HP 8450A diode-array spectrophotometer and a 1-cm path
quartz cell were employed for the absorbance measurements.
À
À
À
NO
2
, NO
3
, and HCOO ions were analyzed by ion chromatography. A Dionex DX-100 ion chromatograph
equipped with an anion self-regenerating suppressor (ASRS-I) and the AI-450 software program was employed
with an Ion Pac AS 14 (Dionex; 250 nm  4 mm) ion-exchange column preceded by an AG 14 guard column
(
Dionex). A 9 mm sodium tetraborate soln. was used as an eluent at an isocratic flow of 1.47 ml/min. Under
À
À
these conditions, the retention times of the above ions were: 2.82 min (HCOO ), 3.65 min (NO
(
2
), and 4.90 min
À
NO
3
).
HCHO was analyzed by spectrophotometry at l ˆ 412 nm, based on the Hantzsch reaction [39].
Total Organic Carbon (TOC) Analysis. A Shimadzu TOC-5050 analyzer was employed for organic ca rbon
determinations, with potassium hydrogen phthalate solns. as TOC standards.
REFERENCES
[
[
1] P. E. Martino, M. I. Diaz Gomez, D. Tamayo, J. Toxicol. Environ. Health 1988, 23, 183.
2] S. Arimoto-Kobayashi, K. Kaji, G. M. A. Sweetmen, P. B. Farmer, H. Hayatsu, Adv. Mass Spectrom. 1998,
4, C096430-1.
1
[
[
3] S. Arimoto-Kobayashi, B. M. Tracey, M. Asao, H. Hayatsu, P. B. Farmer, Mutat. Res. 1999, 444, 413.
4] E. C. Fleming, J. C. Pennington, R. A. Howe, M. R. Colsman, K. E. Garrett, B. G. Wachob, Hazard. Ind.
Wastes 1994, 26, 285.
[
[
[
[
5] Y. L. Chow, Can. J. Chem. 1967, 45, 53.
6] C. H. Bamford, J. Chem. Soc. 1939, 12.
7] C. R. C. Lindley, J. G. Calvert, J. H. Shaw, Chem. Phys. Lett. 1979, 67, 57.
8] C. R. C. Lindley, in −The Kinetics of Dimethylamino Radical Reactions in Simulated Atmospheres: the
Formation of Dimethylnitrosamine and Dimethylnitramine×, The Ohio State University (Dissertation),
1978.
[
9] E. C. Tuazon, W. P. L. Carter, R. Atkinson, A. M. Winer, J. N. Pitts Jr., Environ. Sci. Technol. 1984, 18, 49.
[10] Y. L. Chow, Tetrahedron Lett. 1964, 34, 2333.
[11] J. Polo, Y. L. Chow, J. Natl. Cancer Inst. 1976, 56, 997.
[
[
[
[
[
[
[
[
[
12] Y. L. Chow, M.-P. Lau, R. A. Perry, J. N. S. Tam, Can. J. Chem. 1972, 50, 1044.
13] Y. L. Chow, Acc. Chem. Res. 1973, 6, 354.
14] S. Grilli, M. R. Tost, G. Prodi, Gann 1975, 66, 481.
15] D. Daiber, R. Preussmann, Fresenius Z. Anal. Chem. 1964, 206, 344.
16] T.-F. L. Ho, J. R. Bolton, E. Lipczynska-Kochany, J. Adv. Oxid. Technol. 1996, 1, 170.
17] G. Geiger, J. R. Huber, Helv. Chim. Acta 1981, 64, 989.
18] R.-P. M¸ller, S. Murata, M. Nonella, J. R. Huber, Helv. Chim. Acta 1984, 67, 953.
19] T. Axenrod, G. W. A. Milne, Bull. Chem. Soc. Jpn. 1968, 24, 5775.
20] G. A. Olah, D. J. Donovan, L. K. Keefer, J. Natl. Cancer Inst. 1975, 54, 465.
[21] N. Fiz, J. L. Usero, J. Casado, Int. J. Chem. Kinet. 1993, 25, 341.
[
[
[
22] T. A. Grover, J. A. Ramseyer, L. H. Piette, Free Radical Biol. Med. 1987, 3, 27.
23] D. Grosjean, J. Air Waste Manage. Assoc. 1991, 41, 306.
24] J. N. Pitts Jr., D. Grosjean, K. Van Cauwenberghe, J. P. Schmid, D. R. Fitz, Environ. Sci. Technol. 1978, 12,
946.
[
[
[
25] B. B. Adeleke, J. K. S. Wan, Mol. Photochem. 1974, 6, 329.
26] E. Jakubowski, J. K. S. Wan, Mol. Photochem. 1973, 5, 439.
27] E. A. Kobylinski, B. W. Peterman, in −Evaluation of Ozone Oxidation and UV Degradation of
Dimethylnitrosamine×, Fort Detrick, Frederock, MD 21701, US Army Medical Research and Development
Command, 1979, Report 7908, 1 ± 16.
[
28] D. B. Jobb, R. B. Hunsinger, O. Meresz, V. Taguchi, −Proceedings of the Annual Conference and Exposition
of the AWWA× (June 94) in −Ultraviolet Degradation of N-Nitrosodimethylamine (NDMA) in the
Ohsweken (Six Nations) Water Supply×, New York, NY, 1994, 211.

1
426
Helvetica Chimica Acta ± Vol. 85 (2002)
[29] D. Cooper, L. Keller, −Proceedings of the 10th Technical Seminar on Chemical Spills× (May 93), in
−Advanced Oxidation Technology Demonstration at Ohsweken Six Nations Indian Reserve×, Ottawa,
Canada, 1993, 51.
[
30] J. Polo, Y. L. Chow, IARC Sci. Publ. 1976, 14, 473.
[
31] G. Toussaint, K. W¸rkert, −Mitteilung III der Kommission f¸r Wasserforschung in Verbindung mit der
Kommission zur Pr¸fung von Lebensmittelzusatz- und Inhaltsstoffen×, in −Nitrat ± Nitrit ± Nitrosamine in
Gew‰ssern. D. Toxikologische Bedeutung von Nitrat ± Nitrit ± Nitrosaminen: Meth‰moglobin‰mie im
S‰uglingsalter×, DGF Deutsche Forschungsgemeinschaft, Germany, 1978, p. 136.
32] P. L. Hanst, J. W. Spence, M. Miller, Environ. Sci. Technol. 1977, 11, 403.
33] E. M. Burgess, J. M. Lavanish, Tetrahedron Lett. 1964, 20, 1221.
34] A. L. Fridman, F. M. Mukhametshin, S. S. Novikov, Russ. Chem. Rev. 1971, 40, 34.
35] J. R. Bolton, K. Bircher, W. Tumas, C. A. Tolman, J. Adv. Oxid. Technol. 1996, 1, 13.
36] M. I. Stefan, A. R. Hoy, J. R. Bolton, Environ. Sci. Technol. 1996, 30, 2382.
37] G. Mark, M. N. Schuchmann, H.-P. Schuchmann, C. von Sonntag, J. Water SRT-Aqua 1990, 39, 309.
38] G. J. Hanna, in −Instrumental Methods of OrganicFun tc ional Group Analysis×, Ed. S. Siggia, Wiley-
Interscience, New York, NY, 1972.
[
[
[
[
[
[
[
[
39] T. Nash, Biochem. J. 1953, 55, 416.
Received February 12, 2002