Methods for preparing standard nitrate radical (NO3) gas to calibrate the LIF-based instrument for measurements in the atmosphere

Article abstract of DOI:10.1246/cl.2005.1214

Preparation of standard NO₃ gas is explored at the level of ppbv for atmospheric measurements utilizing an LIF instrument. The sequence of thermal decomposition of N₂O₅ and gas phase titration of NO₃ by adding NO is useful. To reduce NO₂ contamination in N₂O₅, N₂O₅ trapping after mixing NOx and O₃ was adopted. As a convenient method in the field studies, dynamic mixing of NOx and O₃ was explored. Copyright

Full text of DOI:10.1246/cl.2005.1214

1214
Chemistry Letters Vol.34, No.9 (2005)
Methods for Preparing Standard Nitrate Radical (NO3) Gas
to Calibrate the LIF-based Instrument for Measurements in the Atmosphere
Jun Matsumoto,ꢀy Hidekazu Imai, Naohiro Kosugi, and Yoshizumi Kajii
Tokyo Metropolitan University, 1-1 Minami-osawa, Hachioji, Tokyo 192-0397
yDivision of Atmospheric Environment, Solar-Terrestrial Environment Laboratory, Nagoya University,
3-13 Honohara, Toyokawa, Aichi 442-8507
(Received May 10, 2005; CL-050605)
Preparation of standard NO3 gas is explored at the level of
reduce the NO2 impurity, the flow of O3 was added. Then, ac-
quired white powder was used as the N2O5 source. As a result,
measured [NO2] were 8.5 and 11.4 ppmv (10ꢁ6 v/v) before
and after adding NO. [NO3] titrated by NO was determined as
1.5 ppmv. NO2 impurity in the initial N2O5 sample was estimat-
ed as 7.0 ppmv, which was 4.7 times higher than the formed
NO3. Such contamination of NO2 in N2O5 is unavoidable be-
cause of the impurity of NO2 in HNO3, hydrolysis of N2O5,
and thermal decomposition of N2O5 during the exothermic de-
hydration reaction. To reduce the NO2 impurity, the trap-to-trap
purification of N2O5 is essential. However, such a complex pro-
cedure is not suitable for field studies. In addition, it is difficult to
control the yield of N2O5 and the NO3 output at the level of
ppbv.
ppbv for atmospheric measurements utilizing an LIF instrument.
The sequence of thermal decomposition of N2O5 and gas phase
titration of NO3 by adding NO is useful. To reduce NO2 contam-
ination in N2O5, N2O5 trapping after mixing NOx and O3 was
adopted. As a convenient method in the field studies, dynamic
mixing of NOx and O3 was explored.
Nitrate radical (NO3) and dinitrogen pentoxide (N2O5) in
the nocturnal atmosphere have been focused as critical inter-
mediates to discuss NOx cycle, acidification, and atmospheric
lifetime of various compounds.1 Measurements of atmospheric
NO3/N2O5 have been difficult because of their high reactivity
and low abundance at the level of pptv (parts per trillion by vol-
ume; 10ꢁ12 v/v). In this decade, however, a number of observa-
tional studies on NO3/N2O5 were reported.2,3 An instrument for
measuring atmospheric NO3/N2O5 has been developed by the
laser-induced fluorescence technique (LIF).4 The LIF instrument
should be calibrated accurately. Preparation methods for gaseous
NO3 have been reported.5 Halogen atoms can react with nitric
acid (HNO3) to form NO3. However, such radical reactions
are applicable only to the experiments at the low pressure. For
application in the atmosphere, standard NO3 should be supplied
conveniently at the level of ppbv (10ꢁ9 v/v) or less for the
pressure P ¼ 1 atm. For preparing NO3, decomposition of N2O5
is promising.4–6
To synthesize N2O5 controllably, mixing of gaseous NOx
and O3 was adopted. NOx is oxidized by O3 to form NO3 as:
NO þ O3 ! NO2 þ O2;
NO2 þ O3 ! NO3 þ O2:
ð3Þ
ð4Þ
N2O5 is generated from NO2 and NO3 via the reverse reaction of
(1). From reactions (1), (3), and (4), the stoichiometry among
NO, O3, and N2O5 can be described as:
2 NO þ 3 O3 ¼ N2O5 þ 3 O2:
ð5Þ
Excess O3 is needed to rapidly convert NO to N2O5. Figure 1
shows the schematic diagram of experimental configurations
for trapping N2O5 from NOx þ O3 (arrow A) and introduction
of vaporized N2O5 into the LIF NO3/NO2 detector (arrow B).
PFA tubes were used to reduce the loss of NO3/N2O5 on the wall
N2O5 þ M ꢀ NO3 þ NO2 þ M:
ð1Þ
The equilibrium shifts towards NO3 þ NO2 side with tempera-
ture increase. The LIF instrument is capable of monitoring both
NO3 and NO2 when the excitation wavelengths are alternated
between 623 and 613 nm. It is easy to calibrate the instrument
for NO2. Calibration for NO3 is possible when NO3 is converted
to NO2 by NO as:
NO3 þ NO ! 2NO2:
ð2Þ
Increment of NO2 corresponds to twice of decreasing NO3 as
ꢀ½NO2ꢂ ¼ ꢁ2ꢀ½NO3ꢂ. Thus, the information on NO3 concen-
tration can be acquired from the increment of NO2. It is desirable
to preserve stable supplement of gaseous N2O5 with reduced
contamination of NO2. In this study, validation of the methods
for preparing standard NO3 gas is reported to realize an accurate
and/or convenient calibrator for application of the LIF instru-
ment to the field studies.
At first, dehydration of HNO3 by P2O5 was examined, which
is typical for preparing powdery N2O5.7 In a glass flask, 97%
fuming HNO3 was dehydrated by P2O5 for 5 h. Formed N2O5
was gathered at a trap chilled by a methanol bath (183 K). To
Figure 1. Schematic diagram for preparing NO3: (a) LIF instru-
ment; (b) Flow for N2O5 synthesis and supply; (c) In-line supply
for NO3 by mixing of NOx and O3. MFC: mass flow controller;
F1–F6: mass flow rates. Arrows indicate the flows for three
modes: A. N2O5 synthesis, B: supply of N2O5 vapor from the
trap, C: convenient supply of NO3.
Copyright Ó 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.9 (2005)
1215
surface. Representative settings for trapping N2O5 were as fol-
lows: total flow rate F2 þ F3 ¼ 230 sccm (cm3ꢃminꢁ1 at STP);
initial diluted concentrations ½NOꢂ ¼ 0:17%, ½O3ꢂ ¼ 2:9%; re-
action volume Vr ¼ 300 cm3; trap temperature Tt ¼ 195 K
(dry-ice/methanol bath). Because of excess O3, >99% of NOx
can be rapidly converted to N2O5 in the reaction volume. NO2
impurity can be suppressed. The upper limit of N2O5 synthesis
rate was estimated as 5:3 ꢄ 10ꢁ2 gꢃhꢁ1 from the flow rates and
concentrations when complete trapping of N2O5 was assumed.
Note that practical collection efficiency of N2O5 is not 100%
because of heterogeneous loss and incomplete trapping. It is
remarkable that the reactions can be controlled easily only by
adjusting the flow rates and concentrations of initial NOx and
O3. After trapping N2O5, high-grade synthetic air was passed
through the trap and N2O5 was vaporized. N2O5 was diluted to
a suitable level by adding zero air, which was purified by a char-
coal filter. Then, NO2 and NO3 were monitored by the LIF in-
strument. Observed concentrations of NO2 were 209 and 398
ppbv before and after NO addition to the LIF cell, respectively.
Titrated NO3 was 94 ppbv and NO2 impurity in N2O5 was
115 ppbv, which was only 1.2 times higher than the formed
NO3. The LIF instrument can be calibrated for NO3 successfully
with less NO2 contamination. For calibration on N2O5, excess
NO was added just after the N2O5 source to convert N2O5 to
NO2 completely. From reactions (1) and (2), 1 molecule of N2O5
generates 3 NO2. Concentration of N2O5 can be determined
from the NO2 increment. When the heater tube with 1/2 inch di-
ameter and 50 cm length was heated to T ¼ 353 K, the conver-
sion yield of NO3 from N2O5 was 80%.4 The yield was limited
because of the NO3 loss on the wall. After the instrument is cali-
brated, N2O5 concentration can be quantified from the LIF signal
of converted NO3. As another experiment, the stability of the
N2O5 source was explored. After 3 h of reaction and proper di-
lution of N2O5, the averaged concentration of N2O5 was 143 ꢅ
5 ppbv as shown in Figure 2. N2O5 supply was stable for 3 h with
deviation of ꢅ4%. The deviation was mainly due to variation of
bath temperature and consumption of N2O5 powder in the trap.
Short-term variation of N2O5 was <1% as the standard deviation
of 1-min data for 10-min averaging. Consumption rate of N2O5
was estimated from flow rates and observed concentration as
1:2 ꢄ 10ꢁ4 gꢃhꢁ1, which was much smaller than N2O5 synthesis
rate. Thus, reaction period can be shortened in principle. For ex-
ample, it was confirmed that N2O5 was supplied for 7 h at least
after 3 h of reaction. Consequently, the procedure was establish-
ed to calibrate the LIF instrument accurately.
Figure 3. Relationships between calculated [NO3] and LIF
signals for the convenient source of NO3.
out solvent and dry ice. As shown in reactions (3) and (4),
NO3 can be generated when NOx is mixed with excess O3. As
indicated by the arrow C in Figure 1, the convenient supply of
NO3 was constructed. It consisted of a standard NO2 cylinder,
an O3 generator and a mixing system. No wet reagents were used
and the setup was easy and reproducible. The LIF instrument de-
tected the generated NO3 successfully. NO3 concentration can
be calculated theoretically from several parameters (e.g. initial
concentrations, temperature, reaction time, and rate constants),
with gas phase reactions considered.8 For example, ½NO3ꢂ ¼
23 ppbv for F4 ¼ 3000 sccm, F5 ¼ 300 sccm, F6 ¼ 80 sccm,
½NO2ꢂ ¼ 320 ppbv, ½O3ꢂ ¼ 71 ppmv, Vr ¼ 45 cm3, T ¼ 353 K,
and P ¼ 1 atm. The calculation has large uncertainties because
NO3 loss on the wall was not considered. However, the produc-
tion rate of NO3 is reasonable with the gas phase reactions con-
sidered. Thus, the relative response of NO3 on the material NO2
and O3 is proper. Figure 3 shows an example of the relationship
between the observed LIF signal and calculated NO3 concentra-
tion for the convenient source of NO3. As a result, excellent lin-
earity was observed between the signal and the relative concen-
tration (r2 ¼ 0:998). It was confirmed that the convenient source
is promising to prepare the ‘‘relative standard sample’’ of NO3.
From the slope of the regression line, the ‘‘preliminary sensitiv-
ity’’ can be determined in the field as 0.209 cps mWꢁ1 ppbvꢁ1. In
practical uses, after the relative variation of the preliminary
sensitivity is monitored frequently in the field campaign, the
comparison between the convenient source and the accurate
calibration should be explored to determine the correction factor
to acquire the absolute sensitivity. Note that the configurations
and settings of the convenient source should be preserved.
Finally, the secondary standard supply for NO3 was estab-
lished and validated by the LIF instrument. In field studies, it
is desirable to generate the standard sample conveniently with-
References
1
2
R. Atkinson, Atmos. Environ., 34, 2063 (2000).
A. Geyer, B. Alicke, S. Konrad, T. Schmitz, J. Stutz, and U. Platt,
J. Geophys. Res., 106, 8013 (2001).
3
S. S. Brown, H. Stark, T. B. Ryerson, E. J. Williams, D. K. Nicks, Jr.,
M. Trainer, F. C. Fehsenfeld, and A. R. Ravishankara, J. Geophys.
Res., 108, 4299 (2003).
4
5
J. Matsumoto, N. Kosugi, H. Imai, and Y. Kajii, Rev. Sci.
Instrum., 76, 064101 (2005).
R. P. Wayne, I. Barnes, P. Biggs, J. P. Burrows, C. E. Canosa-Mas, J.
Hjorth, G. L. Bras, G. K. Moortgat, D. Perner, G. Poulet, G. Restelli,
and H. Sidebottom, Atmos. Environ., 25A, 1 (1991).
W. R. Simpson, Rev. Sci. Instrum., 74, 3442 (2003).
T. Ishiwata, I. Fujiwara, and I. Tanaka, J. Phys. Chem., 87, 1349
(1983).
6
7
8
IUPAC (2004): http://www.iupac-kinetic.ch.cam.ac.uk/
Figure 2. Long-term stability of N2O5 source.
Published on the web (Advance View) July 30, 2005; DOI 10.1246/cl.2005.1214