Gas-phase reaction of monomethylhydrazine with ozone: Kinetics and oh radical formation

Article abstract of DOI:10.1002/kin.20816

The gas-phase reaction of monomethylhydrazine (CH₃NHi ￡NH₂; MMH) with ozone was investigated in a flow tube at atmospheric pressure and a temperature of 295 ± 2 K using N ₂/O₂ mixtures (3-30 vol% O₂) as the carrier gas. Proton transfer reaction-mass spectrometry (PTR-MS) and long-path FT-IR spectroscopy served as the main analytical techniques. The kinetics of the title reaction was investigated with a relative rate technique yielding k _MMH+O3 = (4.3 ± 1.0) × 10^-15 cm³ molecule^-1 s^-1. Methyldiazene (CH₃Ni ￡NH; MeDia) has been identified as the main product in this reaction system as a result of PTR-MS analysis. The reactivity of MeDia toward ozone was estimated relative to the reaction of MMH with ozone resulting in k _MeDia+O3 = (2.7 ± 1.6) × 10^-15 cm³ molecule^-1 s^-1. OH radicals were followed indirectly by phenol formation from the reaction of OH radicals with benzene. Increasing OH radical yields with increasing MMH conversion have been observed pointing to the importance of secondary processes for OH radical generation. Generally, the detected OH radical yields were definitely smaller than thought so far. The results of this study do not support the mechanism of OH radical formation from the reaction of MMH with ozone as proposed in the literature.

Full text of DOI:10.1002/kin.20816

Gas-Phase Reaction of
Monomethylhydrazine with
Ozone: Kinetics and OH
Radical Formation
S. RICHTERS, T. BERNDT
Leibniz-Institut fu¨r Tropospha¨renforschung e.V., Permoserstr. 15, 04318, Leipzig, Germany
Received 8 April 2013; revised 26 July 2013; accepted 29 July 2013
DOI 10.1002/kin.20816
Published online 3 September 2013 in Wiley Online Library (wileyonlinelibrary.com).
–
ABSTRACT: The gas-phase reaction of monomethylhydrazine (CH3NH NH2; MMH) with ozone
was investigated in a flow tube at atmospheric pressure and a temperature of 295 2 K using
N2/O2 mixtures (3–30 vol% O2) as the carrier gas. Proton transfer reaction–mass spectrometry
(PTR-MS) and long-path FT-IR spectroscopy served as the main analytical techniques. The
kinetics of the title reaction was investigated with a relative rate technique yielding kMMH+O3
=
(4.3 1.0) × 10−15 cm3 molecule−1 s−1. Methyldiazene (CH3N NH; MeDia) has been identified
as the main product in this reaction system as a result of PTR-MS analysis. The reactivity of
MeDia toward ozone was estimated relative to the reaction of MMH with ozone resulting in
=
kMeDia+O3 = (2.7 1.6) × 10−15 cm3 molecule−1
s
−1. OH radicals were followed indirectly by
phenol formation from the reaction of OH radicals with benzene. Increasing OH radical yields
with increasing MMH conversion have been observed pointing to the importance of secondary
processes for OH radical generation. Generally, the detected OH radical yields were definitely
smaller than thought so far. The results of this study do not support the mechanismꢀCof OH
radical formation from the reaction of MMH with ozone as proposed in the literature. 2013
Wiley Periodicals, Inc. Int J Chem Kinet 46: 1–9, 2014
however, is inevitably associated with the formation
of a series of organic oxidation products arising from
the ozonolysis itself and the unwanted, consecutive re-
action of the olefins (used for ozonolysis) with OH
radicals.
A possible application of the reaction of hydrazine
or methyl-substituted hydrazines with ozone as an OH
radical source has been already discussed early in the
1980s [6,7]. Tuazon et al. [6] reported a rate coefficient
of the reaction of hydrazine with ozone at 294–297 K,
kHyd + O3 = 1.4 × 10−16 cm3 molecule−1 s−1, as a result
of a chamber study and concluded that the correspond-
ing rate coefficient for the reaction of monomethylhy-
INTRODUCTION
The OH radical is the most important oxidant in the
atmosphere beside ozone, NO3 radicals, and Cl atoms.
Reliable OH radical sources are needed in labora-
tory experiments investigating atmospheric oxidation
processes. Mostly used approaches are photolysis of
HONO [1], H2O2 [2], O3 [3] or CH3ONO [4]. On
the other hand, ozonolysis of olefins represents a non-
photolytic OH radical source also applicable for low-
NOx studies [5]. This path of OH radical generation,
Correspondence to: T. Berndt; e-mail: berndt@tropos.de
C
ꢀ
–
drazine (CH3NH NH2; MMH) has to be much higher.
2013 Wiley Periodicals, Inc.

2
RICHTERS AND BERNDT
Figure 1 Schematic of the experimental setup.
In contrast to that Martin et al. [8] reported a value of
kMMH + O3 = 8.7 × 10−17 cm3 molecule−1 s−1 mea-
sured at atmospheric pressure in a flow reactor at a
temperature of 299 K. It is noteworthy that in this
experiment a change of the reaction stoichiometry (re-
acted [MMH] per reacted [O3]) was observed start-
ing from unity in the initial stage of the reaction to
a value of up to 7 for rising reactant conversions [8].
From a low-pressure flow tube study at room temper-
ature, however, a rate coefficient kMMH + O3 = 4.5 ×
10−14 cm3 molecule−1 s−1 is reported [9], roughly a
factor of 500 higher compared with the work by Martin
et al. [8].
It is to be noted that OH radical generation has been
shown so far only qualitatively [7]. In the literature,
there is no determination of the OH radical yield given
that confirms the proposed pathways (R1) and (R3).
The aim of this work is to reinvestigate the rate co-
efficient of the reaction of MMH with ozone to over-
come the discrepancy of the kinetics of this step at
room temperature conditions. The experimental deter-
mination of the OH radical yield should help to assess
the title reaction as a potential OH radical source in
laboratory investigations.
The following reaction scheme for the initial step
of the reaction of MMH with ozone is proposed in the
literature [6,8,9]:
EXPERIMENTAL
The experiments were performed in an atmospheric
pressure flow tube with a length of 150 cm and an in-
ner diameter (i.d.) of 15 cm using O2/N2 mixtures (3–
20 vol% O2) as the carrier gas. A total flow was set at
15,000 cm3 min−1 (STP) at 295 2 K leading to a bulk
residence time of 106 s (cf. Fig. 1). Ozone was gen-
erated by flushing 100 cm3 min−1 (STP) gas (10–100
vol% O2 with N2 balance) through an ozone generator
(UVP PS-4), further diluted downstream with O2, and
mixed with the main gas stream (N2, MMH, etc.) in a
turbulent gas mixer just before the entrance of the flow
tube. The turbulent mixing process ensures homoge-
neous reactant concentrations at the flow tube entrance.
MMH was stored in a flask maintained at 228 K, carried
along with 1–20 cm3 min−1 (STP) N2, and diluted im-
mediately with the main N2 stream. Further additions,
such as OH radical scavengers (CO, C3H8, and H2), the
olefins serving as reference substances or C6H6, were
also added to the main N2 stream before it was fed into
the ozone/O2 stream in the turbulent gas mixer. The
needed gas mixtures of 2,3-dimethyl-2-butene (TME),
CH3NH − NH2 + O3 → CH3NH−(N•)H + OH + O2
(R1)
CH3NH−(N•)H + O2 → CH3N = NH + HO2
(R2)
CH3N = NH + O3 → CH3N = N• + OH + O2
(R3)
First, H-abstraction of a nitrogen-bound H-atom by
ozone takes place (R1) followed by an H-abstraction
=
step by O2 forming methyldiazene (CH3N NH; Me-
Dia) (R2). Furthermore, MeDia can also be attacked
by ozone abstracting the remaining nitrogen-bound H-
atom with subsequent OH radical formation (R3). The
pathways (R1)–(R3) describe efficient HOx formation.
As a result of the possible reaction of OH radicals with
MMH and MeDia, secondary products such as dia-
zomethane (CH2N2) and formaldehyde can be formed.
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

GAS-PHASE REACTION OF MONOMETHYLHYDRAZINE WITH OZONE
3
2-methyl-2-butene (MeBut), or benzene were prepared
in a gas-mixing unit. All flow rates were set by cali-
brated mass flow controllers (MKS 1259 and 1179).
The detection of the gas-phase species was carried out
at the outlet of the flow tube by means of long-path
FT-IR spectroscopy (Thermo, Nicolet 6700), proton
transfer reaction–mass spectrometry (PTR-MS; Ioni-
con, HS PTR-QMS 500), online GC-MS analysis (Ag-
ilent, GC 6890 with MSD 5973; separation column:
HP-5MS 19091S-433), and an ozone analyzer (Thermo
Fischer, 49C).
For FT-IR spectroscopy, the whole gas stream was
pumped continuously through a gas cell equipped with
a White mirror system (path length of 20 m). 500–4000
spectra were recorded in the range of 4000–700 cm−1
with a resolution of 1 or 8 cm−1 using a mercury cad-
mium telluride detector. The concentration of phenol
was determined using the calibrated spectrum of the
reference substance measured in this laboratory [10].
MMH was quantified using a cross section of 8.28 ×
10−20 cm2 molecule−1 for the absorption centered at
888 cm−1 [8]. The precision of the MMH concentra-
tion was especially important for the determination of
the OH radical yield. But no error limit of the MMH
cross section was given in the original work. There-
fore, the reliability of the cross section was tested in
experiments with low MMH conversion (<0.3) in the
presence of an OH radical scavenger measuring si-
multaneously the amount of reacted MMH and ozone
(1:1 reaction stoichiometry assumed). Thus, an error
of 20% had to take into account (see the Results and
Discussion section).
In the case of PTR-MS measurements, the gas was
sampled from the center flow at the flow tube outlet to a
first pump stage by means of an uncoated, deactivated
fused silica capillary (0.32 mm i.d.; length 27 cm). A
second capillary (PEEK; 0.25 mm i.d.; length 7 cm)
connected the first pump stage with the drift tube of
the mass spectrometer. The drift tube voltage was set
at 500 V, and the raw data were corrected by the in-
strument’s transmission factor. The PTR-MS operated
in the multiple ion detection mode (MID) or the scan
mode (30–90 amu).
For GC-MS analysis, a small flow was pumped con-
tinuously through a heated GC-loop coupled with a
prefocusing device where the gas sample from the
GC-loop was flushed through an uncoated, deacti-
vated fused silica capillary (0.32 mm i.d.) and trapped
at liquid N2 temperature. After flash heating, the
substances were directly injected at the separation
column.
The initial concentrations were (unit: molecule
cm−3) MMH: (6.6–132) × 1011, ozone: (4.0–780) ×
1011, C6H6: (1.0–1000) × 1012, CO: (1.6–3.2) × 1017,
H2: 8.2 × 1017, C3H8: 1.6 × 1016, TME: 4.1 × 1012,
and MeBut: 4.1 × 1012.
Chemicals
The gases had stated purities as follows: O2 (99.999%;
Air Products, Hattingen, Germany), N2 (99.9999%;
Air Products, further purified with Aeronex gatekeeper
gas purifier SS-500K-I-4R), H2 (99.9992%; Air Prod-
ucts), CO (99.997%; Air Products), C3H8 (99.95%;
Linde, Pullach, Germany), He (99.9997%; Air Prod-
ucts). MMH (98%; Aldrich, Steinheim, Germany),
TME (99.5%; Fluka, Steinheim, Germany), MeBut
(95%; Fluka), and C6H6 (99.9%; Aldrich) and were
used as purchased.
Kinetics
Relative Rate Technique. The relative rate technique
was applied for the determination of the rate coeffi-
cient of the reaction of MMH with ozone. TME and
MeBut served as the reference substances with the cor-
responding rate coefficients kTME+O3 = (1.0 0.2) ×
10−15 cm3 molecule−1 s−1 and kMeBut+O3 = (4.1 0.5)
× 10−15 cm3 molecule−1 s−1 [11]. (Note: Given path-
ways (R1), (R1a), and (R1b) are different only in the
description of the reaction products, but describing the
same reaction with the same rate coefficient k1.)
MMH + O3 → Products
(R1a)
(R4)
Reference + O3 → Products
The measurements were carried out by means of
the PTR-MS in the MID mode monitoring MMH-H+
at m/z = 47 amu, TME-H+ at m/z = 85 amu, and
MeBut-H+ at m/z = 71 amu. These ion traces were
not influenced by any product ions as confirmed in
preexperiments each.
The relative rate coefficient of the reaction of MMH
with O3 (R1a) regarding the reaction of the reference
substance with O3 (R4), k1/k4, was derived according
to Eq. (I):
[MMH]t
[MMH]0
k1 [reference]t
ln
ln
=
(I)
k4 [reference]0
The concentrations in the absence of O3 are ex-
pressed with index “0” and in the presence of O3 with
index “t”. The consumption of MMH and the reference
substance was varied by changing the ozone concen-
tration.
Gas sampling for PTR-MS measurements, GC-MS
analysis, as well as for ozone monitoring was carried
out at the same sampling point.
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

4
RICHTERS AND BERNDT
Consecutive Reaction of MeDia with Ozone. A con-
secutive reaction of MeDia with ozone had to be
taken into account due to the measured concentration
profiles. This reaction path was already discussed in
the literature as a result of former experimental stud-
ies [6,8,9]. The following simple reaction scheme is
assumed allowing the determination of the rate coeffi-
cient ratio k1/k3:
The needed rate coefficient of reactions (R5) and
(R6) were taken from the literature, k5 = 6.11 × 10−11
cm3 molecule−1 s−1 [13] and k6 = 1.26 × 10−12 cm3
molecule−1 s−1 [14]. A phenol formation yield of
0.61
0.07 was applied as reported in the litera-
ture [15]. The phenol concentration and the converted
MMH were measured either by means of FT-IR spec-
troscopy or using PTR-MS analysis.
It is to be noted that only OH radical consumption by
MMH and benzene has been considered for simplicity.
Additional OH radical reactions with MMH products
can lead to an underestimation of y2, especially for
conditions of a relatively high MMH conversion. On
the other hand, additional OH radical generation via the
reaction of the MeDia with ozone can occur resulting in
a possible overestimation of y2. Therefore, the simple
approach used here is able to bring out reliable y2 data
only for conditions of low MMH conversion where
secondary processes are of less importance.
MMH + O3 → y1 · MeDia + · · ·
MeDia + O3 → Products
(R1b)
(R3a)
The resulting differential equations (II) and (III) of
pathways (1b) and (3a)
d [MMH]
= −k1 [O3] [MMH]
(II)
dt
The integrated rate x6 is given by the measured phe-
nol concentration due to the selectivity of phenol for-
mation from reaction (R6):
d [MeDia]
= y1k1 [O3] [MMH] − k3 [O3] [MeDia]
dt
(III)
yield after rearrangement of Eq. (IV)
1
x6 =
[Phenol]t
(V)
0.61
d[MeDia]
d[MMH]
k3 [MeDia]
k1 [MMH]
= −y1 +
(IV)
The value x5 is directly derived from x6 according
to the kinetics of the competitive reactions (R5) and
(R6):
The parameters y1 and k3/k1 were determined us-
ing a least-squares analysis for nonlinear parameter
estimation by means of a damped Gauss–Newton al-
gorithm based on experimental data for MMH and Me-
Dia. The needed integration of Eq. (IV) was done nu-
merically with the help of a semi-implicit step method
[12].
k5 [MMH]
x5 =
x6
(VI)
k6 [benzene]
The benzene concentration is equal to the initial
value due to the very low benzene conversion in the
course of the reaction. In the case of MMH, an av-
eraged concentration is taken, MMH = [MMH]0 –
0.5 · ꢀ[MMH]. The value ꢀ[MMH] stands for the
measured concentration of reacted MMH.
The integrated rate x1 is found from the material
balance for MMH:
OH Radical Formation
The OH radical generation was followed indirectly in-
vestigating the phenol formation from the reaction of
OH radicals with benzene being in competition with
the reaction of OH radicals with MMH (cf. pathways
(R5) and (R6)). The knowledge of the integrated rates
x1, x5, and x6 according to the reactions (R1c), (R5),
and (R6), respectively, and their rate coefficients k5 and
k6 allows a rough estimate of the OH radical yield y2
from reaction (R1c):
ꢀ
ꢁ
x1 = ꢀ[MMH] − x5
(VII)
O3 + MMH → y2 · OH + · · ·
OH + MMH → Products
(R1c)
(R5)
(R6)
Setting the OH radical formation equal to the loss
processes, Eq. (VIII) follows:
x5 + x6
y2 =
(VIII)
x1
OH + C6H6 (+O2) → 0.61 · C6H5OH + · · ·
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

GAS-PHASE REACTION OF MONOMETHYLHYDRAZINE WITH OZONE
5
Figure 2 Typical difference spectrum from PTR-MS
Figure 3 Product formation at 45 amu (MeDia) and 43 amu
(CH2N2) as a function of reacted MMH, PTR-MS analysis.
Initial concentrations (molecule cm−3): [MMH] = 6.6 ×
1011; [O3] = (4.0–250) × 1010; [H2] = 8.2 × 1017; [O2]
= 8.2 × 1017. The dashed line shows the modeling results
according to Eq. (IV).
analysis (ozone production: on/off). Initial concentrations
(molecule cm−3): [MMH] = 6.6 × 1011; [O3] = 5.0 × 1012
;
[H2] = 1.6 × 1017; [O2] = 8.2 × 1017
.
RESULTS AND DISCUSSION
Product formation
spectroscopy at 1628 cm−1. NO was transformed to
NO2 under the chosen experimental conditions of high
ozone concentrations. The found NO2 concentration
was below the detection limit of the FT-IR analysis
(5 × 109 molecule cm−3) even in experiments with the
highest MMH conversion of about 1 × 1013 molecule
cm−3. A NOx yield <0.0005 can be inferred from
these measurements. It was also tested for possible
HCHO formation using FT-IR measurements with a
resolution of 1 cm−1. The expected signal centered at
1746 cm−1 (a detection limit of about 1010 molecule
cm−3) was only slightly above the background scat-
tering, and the reproducibility of this signal from one
experiment to the next was insufficient. A conservative
estimate results in a HCHO yield <0.005.
In Fig. 2, a typical PTR-MS difference spectrum (ozone
production: on/off) from a reaction in the presence of
H2 as an OH radical scavenger is shown. In this run,
ozone was chosen in large excess over MMH resulting
in a MMH conversion >0.9. Reacted MMH appears as
a negative signal at 47 amu (MMH-H+) and the prod-
ucts as positive signals. The main product at 45 amu
was assigned to protonated MeDia. The formation of
MeDia is in line with observations from former stud-
ies [6,8,9]. The signal at 43 amu was tentatively as-
signed to protonated diazomethane. This assignment
was supported by simultaneously performed FT-IR
measurements showing a distinct absorption centered
at 2115 cm−1, typical for diazomethane [6]. At the mo-
ment, there exists no reasonable explanation regarding
the chemical composition of other products responsi-
ble for the signals at 49, 59, and 75 amu. In Fig. 3 the
product signal strength at 45 amu (MeDia) and 43 amu
(CH2N2) is depicted as a function of converted MMH
from a measurement series in the presence of H2 for
OH radical scavenging. A curvature in the plot of Me-
Dia (45 amu) versus reacted MMH is visible, especially
for a relative high MMH conversion. This fact points
to a consecutive reaction of MeDia, most probably by
ozone (cf. pathway (R3)). On the other hand, the shape
of the curve of the signal at 43 amu attributed to CH2N2
indicates that this substance is produced in secondary
processes, at least partly.
Kinetics of the Reaction of MMH with
Ozone
The relative depletion of MMH (47 amu) and reference
substance (here TME at 85 amu) in the reaction with
ozone (in the presence of OH radical scavengers) was
followed by means of PTR-MS measurements using
the MID (cf. Fig. 4). Increasing reactant consumption
(MMH and the reference substances) was achieved by
increasing ozone concentrations. After switching on
the ozone generator, rising ozone production occurs
within the warm-up period of 5–10 min providing the
desired change of the ozone concentration. The mea-
sured data for MMH conversions ≤0.6 were taken for
Furthermore, it was tested for possible NOx for-
mation using NO2 measurements by means of FT-IR
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

6
RICHTERS AND BERNDT
Figure 4 Reaction of MMH (47 amu) and TME (85 amu)
with ozone, PTR-MS analysis. Initial concentrations
(molecule cm3): [MMH] = 6.6 × 1011; [O3]max = 2.5 ×
1012; [TME] = 4.1 × 1012; [CO] = 1.6 × 1017; [O2] =
8.2 × 1017
.
the kinetic analysis according to Eq. (I) (see Fig. 5).
The experiments were carried out with two reference
substances, TME or MeBut, two different OH radi-
cal scavengers, C3H8 or CO, and for two different O2
concentrations balances by N2, 8.2 × 1017 or 4.9 ×
1018 molecule cm−3. No significant influence of one of
these parameters on the resulting rate coefficient was
obtained. The experimental conditions along with the
kinetic findings from a linear regression analysis are
summarized in Table I. All errors given represent 2σ
Figure 5 Experimental data for the reaction of MMH with
ozone plotted according to Eq. (I). Experimental conditions
and the findings are summarized in Table I. The lines rep-
resent the individual result from the regression analysis for
each measurement series.
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

GAS-PHASE REACTION OF MONOMETHYLHYDRAZINE WITH OZONE
7
Table II Results of the Sensitivity Study Investigating
the Effect of the PTR-MS Detection Efficiency for MMH
and MeDia on the Parameters k3/k1 and the Yield y1
limits. In the case of the rate coefficient k1, the propa-
gation of error includes the experimental error of k1/k4
and the error of the rate coefficient of the reference
reaction.
k(MMH + H3O+)/
k(MeDia + H3O+)
k3/k1
y1
As a result of this study, a rate coefficient kMMH+O3
= (4.3 1.0) × 10−15 cm3 molecule−1 s−1 at 295
2 K can be given. This value is qualitatively in line with
the finding by Tuazon et al. [6] stating a rate coefficient
for kMMH+O3 to be much higher than 1.4 × 10−16 cm3
0.5
1.0
2.0
0.63 0.34
0.63 0.34
0.63 0.34
0.27 0.02
0.54 0.04
1.07 0.09
molecule−1 −1. However, our value is neither in a
s
reasonable agreement with the data by Martin et al.
(kMMH + O3 = 8.7 × 10−17 cm3 molecule−1
s
−1) [8]
the reaction of MeDia with ozone k3 = (2.7
1.6)
nor with that by Coleman et al. (kMMH + O3 = 4.5 ×
10−14 cm3 molecule−1 s−1) [9]. An explanation for this
distinct disagreement of our value with both literature
data cannot be specified.
× 10−15 cm3 molecule−1 s−1 follows. The given error
includes the uncertainty of k3/k1 and the error of the
rate coefficient of the reference reaction (2σ limits).
Martin et al. [8] reported a rate coefficient for the
reaction of MeDia with ozone kMeDia+O3 = (2.5 0.1)
× 10−16 cm3 molecule−1 s−1 about one order of mag-
nitude lower than our value. Also here, no reasoned
explanation for the discrepancy can be given.
Kinetics of the Consecutive Reaction of
MeDia with Ozone
The product analysis revealed that the formed Me-
Dia undergoes a consecutive reaction most likely with
ozone (in the presence of an OH radical scavenger)
(cf. Fig. 3. This pathway was already proposed in the
literature [6,8,9]. The rate coefficient of the reaction
of MeDia with ozone, k3, was estimated relative to k1
based on the MeDia measurements as a function of re-
acted MMH as given in Fig. 3, [MMH] = 6.6 × 1011;
[H2] = 8.2 × 1017; [O3] = (4.0–250) × 1010 molecule
cm−3. It was assumed that the signal measured at
45 amu from the PTR-MS analysis was exclusively due
to MeDia. According to reactions (R1b) and (R3a), the
differential equation (IV) follows. Equation (IV) was
integrated numerically. Supposing first of all that the
efficiency of PTR-MS detection for MMH and MeDia
is equal (i.e., the same rate coefficients of protonaton,
k(MMH + H3O+)/k(MeDia + H3O+) = 1), the mea-
sured signals of MMH and MeDia can be used for the
nonlinear parameter estimation without any correction
factor. The initial values of MMH and MeDia needed
for integration are set at 7450 and 0 cps, respectively,
according to the measured signal strengths at 47 and
45 amu without ozone addition. The nonlinear param-
OH Radical Formation
Formed OH radicals from the reaction of MMH with
ozone were detected indirectly using phenol forma-
tion from the reaction of OH radicals with benzene.
Measurements with all analytical techniques (GC-MS,
PTR-MS, FT-IR) confirmed the formation of phenol
in a wide range of reactant concentrations; [MMH] =
6.6 × 1011 and [C6H6] = 5.0 × 1012 molecule cm−3
used for PTR-MS and GC-MS analysis; [MMH] =
1.3 × 1013 and [C6H6] = 1.1 × 1015 molecule cm−3
used for FT-IR and GC-MS analysis. The O2 concen-
tration was set at 8.2 × 1017 or 8.2 × 1018 molecule
cm−3, and ozone was varied to achieve different MMH
conversions.
The diffusion-controlled wall loss of OH radicals
was estimated considering the first-order rate law with
kwall = 3.65·D/r2 = 0.014 s−1 using D = 0.22 cm2 s−1
[16]. Even for the lowest benzene concentration of
5.0 × 1012 molecule cm−3, k6·[C6H6] = 6.3 s−1 follows
making the wall loss of OH radicals negligible under
all conditions.
eter estimation yielded y1 = 0.54
0.04 and k3/k1
= 0.63 0.34. The dashed line in Fig. 3 represents
the result of the fitting procedure. A sensitivity study
showed that the obtained ratio k3/k1 was independent
of the assumed PTR-MS detection efficiency for MMH
and MeDia. The parameter y1 (MeDia yield from re-
action (R1b)), however, showed a clear dependence on
this assumption. Table II summarizes the results of the
sensitivity study assuming k(MMH + H3O+)/k(MeDia
+ H3O+) = 0.5 or 2. Using our value k1 = (4.3 1.0)
For instance, Fig. 6 shows results from FT-IR mea-
surements comparing the detected reaction products
from runs in the presence (a, b) and the absence (c) of
benzene. Phenol formation has been identified by its
absorptions at 1607 and 3652 cm−1 (a). After phenol
subtraction using a reference spectrum, there was re-
maining product absorption at 1714 cm−1 arising from
benzene oxidation (b). This signal is very likely due to
unsaturated carbonyls, known products of benzene ox-
idation beside phenol [10]. No effort was undertaken in
× 10−15 cm3 molecule−1
s
−1, a rate coefficient for
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

8
RICHTERS AND BERNDT
MMH concentration, the initial reaction stoichiome-
try (reacted [O3] per reacted [MMH]) for low MMH
conversion (smaller than 0.3) was measured yielding
ꢀ[O3]/ꢀ[MMH] = 0.85 0.08. This points to an er-
ror of the MMH determination of <20% assuming that
the deviation from the expected value of unity was
mainly due to uncertainties of the MMH concentra-
tions. An error of the MMH concentration of 20% has
been taken into account for the calculation of the OH
radical yield. The ratio ꢀ[O3]/ꢀ[MMH] increased sig-
nificantly for higher MMH conversions obviously ac-
cording to secondary processes consuming additional
ozone (consecutive MeDia reaction, etc.).
In Fig. 7, the findings for the OH radical yield as a
function of converted MMH are depicted. In the error
bars, the uncertainties of the concentration determina-
tion and the errors of the rate coefficients needed in the
calculation are considered. The obtained OH radical
yields are substantially lower than the expected yield
of unity as given in the literature according to reaction
(R1). Moreover, the increase in the OH radical yield
with increasing MMH conversion indicates the impor-
tance of secondary processes for OH radical formation
rather than substantial OH radical production from the
reaction of MMH with ozone (R1). The shape of the
OH radical yield as a function of reacted MMH in
Fig. 7 is similar to the shape of the curve describing
the MeDia production as given in Fig. 3. But the used
approach (Eqs. (V)–(VIII)) is strictly speaking only
valid for low MMH conversions (see the Experimen-
tal section), i.e., negligible importance of secondary
processes. So we cannot draw any reliable conclusion
regarding the OH radical yield from the reaction of
MMH with ozone (R1). It could be speculated from
Figure 6 FT-IR spectra (resolution: 8 cm−1, 2000 scans) of
the products from the reaction of MMH with ozone, residual
reactants have been subtracted: (a) and (b) Product formation
in the presence of benzene; in (b) absorptions from phenol are
subtracted. (c) Product formation in the absence of benzene.
Absorptions at 1607 and 3652 cm−1 are due to phenol, at
1714 cm−1 from an unsaturated carbonyl of the benzene
oxidation, at 2115 cm−1 assigned to CH2N2, and at 2963
cm−1 from a MMH oxidation product not specified. Initial
concentrations (molecule cm−3): [MMH]0 = 1.3 × 1013
,
[O3]0 = 7.8 × 1012, and [benzene]0 = 1.1 × 1015
.
the present study to identify the carbonylic substances
formed from benzene.
An attempt was undertaken to quantify the OH rad-
ical production in this system. Within a simple re-
action scheme, pathways (R1c), (R5), and (R6), the
OH radical formation yield from the reaction of MMH
with ozone was estimated with the help of Eqs. (V)–
(VIII). FT-IR analysis (high reactant concentrations)
and PTR-MS measurements (low concentration) were
applied to obtain the needed concentrations of phenol
and converted MMH. For FT-IR analysis, a calibrated
phenol spectrum from our laboratory was used and for
the determination of the MMH concentration a cross
section of 8.28 × 10−20 cm2 molecule−1 for the ab-
sorption at 888 cm−1 as given in the literature [8].
In the case of PTR-MS measurements, the MMH-H+
signal at 47 amu was calibrated against simultaneously
performed FT-IR measurements at 888 cm−1 for ele-
vated MMH concentrations. The C6H5OH-H+ signal
at 95 amu was calibrated against the 79 amu signal of
a well-defined benzene concentration with respect to
the different protonation rate coefficients k(C6H5OH
+ H3O+)/k(C6H6 + H3O+) = 1.26 [17]. It is to be
noted that our determination of the MMH concentra-
tion is merely based on a single cross section from the
literature [8]. To evaluate the precision of the absolute
Figure 7 Derived OH radical yields from the reaction of
MMH with ozone according to a simplified reaction scheme,
pathways (R1c), (R5), and (R6).
International Journal of Chemical Kinetics DOI 10.1002/kin.20816

GAS-PHASE REACTION OF MONOMETHYLHYDRAZINE WITH OZONE
9
our experiments that the OH radical yield from (R1)
is very low or close to zero. Nevertheless, OH radicals
are clearly formed in the overall process, but especially
by secondary processes.
ratory regarding the OH radical–initiated oxidation of
isoprene under low-NOx conditions were promising.
The authors thank R. Graefe, K. Pielok, and A. Rohmer for
technical assistance.
CONCLUSIONS
BIBLIOGRAPHY
The gas-phase reaction of MMH with ozone has been
studied in a flow tube at atmospheric pressure and a
1. Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L.
P. J Phys Chem 1978, 82, 132.
2. Ohta, T. J Phys Chem 1983, 87, 1209.
3. Robertson, R.; Smith, G. P. J Phys Chem A 2006, 110,
6673.
4. Atkinson, R.; Aschmann, S. M.; Winer, A. M.; Pitts, J.
N. Int J Chem Kinet 1982, 14, 507.
5. Siese, M.; Becker, K. H.; Brockmann, K. J.; Geiger, H.;
Hofzumahaus, A.; Holland, F.; Mihelcic, D.; Wirtz, K.
Environ Sci Technol 2001, 35, 4660.
6. Tuazon, E. C.; Carter, W. P. L.; Winer, A. M.; Pitts, J.
N. Environ Sci Technol 1981, 15, 823.
7. Tuazon, E. C.; Carter, W. P. L.; Atkinson, R.; Pitts, J. N.
Int J Chem Kinet 1983, 15, 619.
8. Martin, N. B.; Davis, D. D.; Kilduff, J. E.; Mahone, W.
C. Final Report, ESL-TR-89–32, Air Force Engineer-
ing and Services Center, Tyndall Air Force Base, Bay
County, near Panama City, FL, 1989.
9. Coleman, D. J.; Judeikis, H. S.; Lang, V. Final Report,
AL/EQ-TR-1993–0029, Air Force Material Command,
Tyndall Air Force Base, Bay County, near Panama City,
FL, 1996.
temperature of 295
2 K. MeDia was identified as
the main product as a result of the PTR-MS analy-
sis. The kinetics of the reaction of MMH with ozone
was investigated with a relative rate technique yielding
kMMH+O3 = (4.3 1.0) × 10−15 cm3 molecule−1 s−1
.
The rate coefficient of the consecutive reaction of Me-
Dia with ozone was estimated relative to the reaction
of MMH with ozone resulting in kMeDia+O3 = (2.7
1.6) × 10−15 cm3 molecule−1
s
−1. OH radicals were
measured indirectly by following the phenol formation
from the reaction of OH radicals with benzene. Mea-
sured OH radical yields were substantially lower than
expected from the pathway describing OH radical gen-
eration from MMH + O3 with a yield of unity (R1) as
given in the literature. The observed increase in the OH
radical yield with increasing MMH conversion points
rather to the importance of secondary processes for OH
radical generation.
NOx formation was below the detection limit of the
FT-IR analysis, yielding a NOx yield <0.0005 for the
highest MMH conversion of about 1 × 1013 molecule
cm−3. There was no clear FT-IR spectroscopic evi-
dence for the production of HCHO (an imaginable
product of the methyl group). The HCHO signal from
FT-IR measurements was only slightly above the back-
ground scattering, and the reproducibility from one ex-
periment to the next was poor. A conservative estimate
results in a HCHO yield <0.005.
10. Berndt, T.; Bo¨ge, O. Phys Chem Chem Phys 2001, 3,
4946.
11. Witter, M.; Berndt, T.; Bo¨ge, O.; Stratmann, F.;
Heintzenberg, J. Int J Chem Kinet 2002, 34, 394.
12. Stoer, J.; Bulirsch, R. Einfu¨hrung in die Numerische
Mathematik II; Springer-Verlag: Berlin, 1987 (in
German).
13. Vaghjiani, G. L. Int J Chem Kinet 2001, 33, 354.
14. Atkinson, R. Chem Rev 1986, 86, 69.
15. Berndt, T.; Bo¨ge, O. Phys Chem Chem Phys 2006, 8,
1205.
Although the reaction of MMH with ozone is a less
efficient OH radical source than thought, it represents a
useful nonphotolytic way of OH radical production for
low-NOx experiments. First experiments in our labo-
16. Ivanov, A. V.; Trakhtenberg, S.; Bertram, A. K.; Ger-
shenzon, Y. M.; Molina, M. J. J Phys Chem A 2007,
111, 1632.
17. Ionicon, HS. PTR-MS Manual; Ionicon Analytic, 2010.
International Journal of Chemical Kinetics DOI 10.1002/kin.20816