| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Pediatrics, Indiana University, Indianapolis, IN 46202.
2
Department of Pediatrics, Stanford University Medical
Center, Stanford, CA 94304.
3
Neonatal Network, National Institute of Child Health and
Human Development, Bethesda, MD 20892.
4
Department of Biomedical Engineering, Children's
Hospital National Medical Center, Washington, DC 20010.
5
Analytical Chemistry Division, National Institute of
Standards and Technology, Gaithersburg, MD 20899.
6
Identification of commercial equipment in this article is to provide as complete a description of the experimental procedures followed as possible. Such identification does not constitute a recommendation or endorsement by NIST of the vendor or equipment.
a Address correspondence to this author at: Division of Neonatal-Perinatal Medicine, Indiana University School of Medicine, 702 Barnhill Drive, Room RR-208, Indianapolis, IN 46202. Fax 317-274-2065; e-mail gsokol{at}iupui.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: The formation of NO2 was studied under dynamic flows emulating a base case NO ventilator mixture containing 80 ppm NO in a 90% oxygen matrix. The difficulty in measuring NO2 concentrations below 2 ppm accurately was overcome by the use of tunable diode laser absorption spectroscopy.
Results: Using a second-order model, the rate constant, k, for NO2 formation was determined to be (1.19 ± 0.11) x 10-11 ppm-2s-1, which is in basic agreement with evaluated data from atmospheric literature.
Conclusions: Inhaled NO can be delivered safely in a well-designed, continuous flow neonatal ventilatory circuit, and NO2 formation can be calculated reliably using the rate constant and circuit dwell time. © 1999 American Association for Clinical Chemistry
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
80 ppm appear to be without direct toxic effects when inhaled for
short periods of time (1)(2)(3)(4)(5)(6)(7).
When NO is mixed with oxygen in a gaseous environment, nitrogen
dioxide (NO2) is formed spontaneously by a
termolecular mechanism (8). The chemical reaction is
described in Eq. 1
:
 | (1) |
The rate expression for the formation of NO2
is given by Eq. 2 , which shows this reaction to have first-order
dependence on oxygen concentration and second-order dependence on NO
concentration:
 | (2) |
NO2 is a toxic gas, and the Occupational Safety and Health Administration limits human peak exposure to 5 ppm (9). However, alterations of airway reactivity have been reported in humans at exposures as low as 1.5 ppm (10).
Concern has focused on delivering inhaled NO in a high oxygen concentration environment. The amount of NO2 formed spontaneously and subsequently inspired by the patient is of prime importance. Several investigators reported their experience regarding NO2 formation during inhaled NO delivery (11)(12)(13)(14)(15)(16). The rates of formation of NO2 were calculated or measured, but varied appreciably between reports. A secondary issue is the ability of presently available analyzer systems to measure NO2 accurately in this environment (15)(17).
The goal of this study was to examine the rate of NO2 formation under clinical conditions in a continuous flow neonatal ventilator circuit. The measurement of NO2 was accomplished by the use of a tunable diode laser absorption spectrometer (TDLAS), which permitted the accurate measurement of NO2 without interference from components of the inhaled NO ventilator matrix.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, circuit A, site A). Gas was sampled for analysis by the TDLAS
vacuum inlet immediately proximal to the endotracheal tube adapter
(Fig. 1
, circuit A, site B). The volume of the circuit through which
the inspired gas mixture flowed after NO and oxygen were mixed and
which ended at the patient (inlet of TDLAS) was 0.193 ± 0.005 L.
The elapsed time for the NO in oxygen mixture to reach the TDLAS is the
system dwell time. The ventilator was used to study
NO2 formation for dwell times between 1.37 and
5.57 s.
|
laboratory gas mixing circuit
To measure NO2 formation at longer dwell
times, a special Pyrex dual chamber gas mixing apparatus was used to
provide a circuit volume of 0.653 ± 0.003 L (Fig. 1
, circuit B).
The mixing apparatus was used to study NO2
formation for dwell times between 3.9 and 13 s.
Experiments were conducted at ambient conditions with a room temperature of 21.0 ± 0.5 °C at a circuit pressure of 0.101 ± 0.002 MPa (1 atmosphere). Twenty-two experiments covering 13 dwell times were conducted over a 3-day period, with 7 dwell time conditions being repeated on different days. NIST gaseous Standard Reference Material 2735, which is certified to contain 793.5 ± 4.0 ppm NO and 1.9 ± 0.2 ppm NO2, was used as the NO source gas. For each experiment, Standard Reference Material 2735 was dynamically diluted 1:9 with 99.6% purity oxygen, using calibrated mass flow controllers (MFCs) to give NO concentrations between 79 and 81 ppm. The total gas flow rate was varied between 3 and 10 L/min through the ventilator circuit or the special mixing apparatus to achieve 13 dwell times to investigate NO2 formation. The mass flows were allowed to equilibrate for a minimum of 5 min, after which the amount of NO being delivered was verified by use of a chemiluminescence monitor (CLD 700 AL; Eco Physics). The NO2 concentration was determined from the ratio of an integrated NO2 TDLAS absorbance line profile calibrated against a NIST gas standard. The TDLAS 30-L gas cell contained 72 kPa-L of sample under dynamic flow at approximately 52 kPa-L per minute. This translates to an average dwell time in the TDLAS cell of 1.4 min, during which some NO2 will form even at the greatly reduced pressure of 2.4 kPa. The total cell background for the experiments each day was estimated by a linear regression plot of the inverse of the total circuit flow rate (x values) vs the measured TDLAS NO2 output (y values). The y-intercept of the linear plot represented the total NO2 background as the circuit flow rate approaches infinity (x = 0). The cell contribution to the total background value was estimated by subtracting the NO2 impurity of the diluted NO source mixture (Standard Reference Material 2735) from the total intercept value.
gas flow measurement
The MFCs (Model FC260; Tylan Corp.) were calibrated with a
volumetric gas flow calibrator (Califlow A-100; MKS Instruments) for
the NO source mixture and the dilution oxygen. The equation for the
mass flow calibration line for each flowmeter was used to predict the
MFC readings necessary to produce 1:9 dilution ratios for total flows
of 3–10 L/min.
tdlas set-up
The NIST TDLAS system has been described in detail previously
(18). Briefly, the sample gas is flowed continuously through
a White-type gas cell estimated to have a volume of 30 L. The inlet
flow rate at the cell is adjusted as necessary to maintain a cell
pressure of 2.40 ± 0.07 kPa (18 Torr). The inlet flow rate is
estimated to be ~0.5 L/min. The cell pressure is monitored by a
calibrated capacitance manometer. The path length of the multipass cell
was set to 81.59 m. A lead salt diode laser (Laser Photonics) with a
tuning range centered around 2940 cm-1 was set
to scan over a single ro-vibrational transition of the
NO2
1+3 band
(19). The narrow tuning range of the laser minimized any
interference from species such as water in the system. The
second-derivative absorbance signal was monitored by modulating the
frequency of the diode laser at 36.3 KHz and processing the signal with
a high-speed lock-in amplifier.
The line strengths for the NO2 transitions in the
2940 cm-1 region are on the order of 1 x
10-23
cm-1/(molecules/cm2)
(20). For the experimental conditions described above, the
absorbances [ln(Io/I)]
are approximately 0.005 or less, where
Io and I are the incident
and transmitted intensities. At this absorbance, the sample is
optically thin, and the peak-to-peak second harmonic signal at the line
center varies linearly as a function of the NO2
concentration and is given by the product:
 | (3) |
where L is the path length, P is the sample
pressure, 
is the absorption coefficient at line
center, and H2 accounts for the
modulation effects (21). The NO2
concentration was obtained by comparing the signal strength to the
signal obtained from a NIST gas standard containing 4.5 ± 0.2 ppm
NO2 in nitrogen.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
summarizes the measured data for determining the
NO2 formation rate constant. The rate constant
for NO2 formation, k, was determined
to be 1.19 ± 0.11 x 10-11
ppm-2 s-1. Alternately,
k can be expressed as 1.64 ± 0.16 x
10-38
(molecules/cm3)-2
s-1 or 6.00 ± 0.57 x
103 (mol/L)-2
s-1. The relative SD for the 13 measurements of
k was ± 2.7%. This imprecision is part of the ±
9.5% total uncertainty of the k value as explained below in
the "Discussion of Errors" section. Plugging k into Eq. 2
leads to the rate expression for NO2 formation:
 |
 | (4) |
|
The current study focuses on NO2 formation
at the clinically important conditions of a high (90%) oxygen matrix
at ambient pressure and temperature. An extensive body of previous
NO2 formation studies exists in support of
atmospheric chemical processes, most of which describe
NO2 formation experiments at low pressures or in
air. In 1992, Atkinson et al. (22) critically reevaluated
the published kinetic data for the formation of
NO2 from NO and O2 as part
of a general review of relevant kinetic expressions used for
atmospheric modeling. Atkinson et al. gave a "preferred expression"
from which the rate constant k was calculated as equal to
1.45 ± 0.22 x 10-11
ppm-2 s-1 for 101 kPa (1
atmosphere) and 21 °C (see Table 3
). This value agrees within
± 20% with the measured k above, falling within their
respective measurement uncertainties. The data presented here show that
in all likelihood the kinetics of NO2 formation
are the same whether describing inhaled NO therapy or atmospheric smog
reactions. For example, at the clinically important conditions of 80
ppm NO in a 90% oxygen matrix, Eq. 4
gives:
 |
 | (5) |
|
 | (6) |
Thus, NO2 is predicted to form at a rate of
~0.14 ppm per second of dwell time for 80 ppm NO in a 90% oxygen
matrix. Fig. 2
is a plot of 13 measured data points taken from Table 1
, which
shows the linear dependence of the NO2 formed as
a function of dwell time.
|
discussion of uncertainty
The estimates of measurement uncertainty are reported as expanded
values (2 
), using U =
kU%RSD, where the expansion factor
k = 2 and U%RSD is the
uncertainty in a measurement expressed as a percentage of relative SD.
The expanded uncertainty in the NO concentration estimated from the
mass flow measurements is 2%. The uncertainty in the system dwell time
is 3%. The measurement of NO2 on the TDLAS is
limited by ambient variations in the cell temperature (0.4%),
variations in the cell pressure (3%), and the TDLAS absorbance
integration imprecision (2%). The uncertainty because of systematic
effects in the NO2 reference standard (4.4%) is
combined with the random uncertainty. The overall uncertainty in the
measured rate constant is given by:
 | (7) |
 | (8) |
As discussed above, the imprecision from 13 measured values of the
rate constant shown in Table 1
is 2.7%RSD, which when expanded by the
appropriate t-value for the 95% confidence interval
(t = 2.2) is equal to 5.9% (2 ). This value is
consistent with the ± 5.1% value calculated under quadrature in
Eq. 8
. The relative expanded uncertainty, UK, in
the rate constant is ± 9.5% relative, and the rate constant is
expected to lie within ± UK with a level
of confidence close to 95% (23).
no2 formation dependence on no concentration
Additional experiments were carried out to demonstrate the
second-order dependence of the rate of formation of
NO2 on the NO concentration. Table 2
shows the NO2 concentration measured by
TDLAS for NO values ranging from 10 to 100 ppm at a constant dwell time
of 3.91 s. The measured NO2 concentrations
are compared in Table 2
with calculated values for
NO2 formation predicted using Eq. 4
and the
"preferred equation" from Atkinson et al. (22).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) and one rate constant is needed to
model NO2 formation in a patient delivery circuit
under a broad range of test conditions. The rate constant k
was measured here to be 1.19 ± 0.11 x
10-11 ppm-2
s-1 and was reasonably estimated (± 20%) from
the evaluated rate constant given by Atkinson et al. (22),
adjusted for 101 kPa (1 atmosphere) and 21 °C (see Table 3
).
Some clinical studies have raised confusion when describing measured
NO2 formation using other formats, including
reaction models, kinetic rate expressions, and rate constants. One
example, Nishimura et al. (16), provided a comprehensive
study that investigated NO2 formation during a
simulation of inhaled NO therapy in adults. Nishimura et al. used an
integrated form of Eq. 9 , published by Glasson and Tuesday
(25) in 1963:
 | (9) |
Although Nishimura et al. (16) and this study are in basic agreement on the concentrations of NO2 formed from similar experiments, their reported rate constants are a factor of 2 greater than those reported by Atkinson et al. (22) and this study. Much of the "confusion" can be eliminated by presenting results in a common format and by using reasonable estimates of ventilator circuit dwell times.
The clinical environment of 80 ppm NO in a 90% oxygen matrix was
selected for investigation because it represents the high end of NO
concentrations under use in clinical trials, and therefore the maximum
rate of NO2 formation likely to be encountered.
When NO is being delivered in this environment,
NO2 formation is ~0.14 ppm/s. The recommended
ventilator circuit that adds the NO source gas as close to the patient
as practical (Fig. 1
, circuit A, site A) will have a small system
volume (<0.1 L). The use of recommended total flow rates exceeding 10
L/min (0.167 L/s) will limit the dwell time to <0.6 s. When Eq. 6
is
used, the maximum NO2 formed will be 0.08
ppm. It has been verified in this study that when delivering
ventilator gases with NO concentrations<80 ppm, the rate of
NO2 formation will decrease as a function of the
square of the NO concentration. This dramatic lowering of
NO2 formation is predicted by Eq. 4
and
demonstrated by measured data presented in Table 2
.
effect of humidity
Although the majority of the measurements were performed in a dry
circuit at 21 °C, two experiments were performed using the Sechrist
ventilator where humidity was added to the dilution oxygen. The
addition of the hot saturated water vapor heated the gas in the
ventilator circuit to 36 °C. The observed concentration of
NO2 formed (TDLAS) was observed to be ~20%
lower in the humidified gas. This decrease is consistent with the
termolecular mechanism (Eq. 1
) where NO2
production has previously been observed to decrease continuously with
increasing temperature until a minimum is reached at 327 °C
(8). Thus, the observed effect on the
NO2 production after humidity was added to the
circuit was consistent with the expected rate decrease predicted by the
higher temperature of the gas in the delivery circuit.
other no2 sources
This investigation resolves the question of
NO2 formation within a continuous flow delivery
circuit. However, there are additional sources of
NO2 that can be delivered to the patient. The NO
source cylinder always contains a small concentration of
NO2, which will lead to the delivery of some
NO2 to the patient. For example, an 800-ppm NO
source cylinder often contains 8 ppm NO2, which
is only 1% contamination relative to the NO concentration. This source
gas is then diluted 1:9 to achieve a ventilator mixture containing 80
ppm NO and 0.8 ppm NO2 that is being delivered to
the patient. In this example, there is 10-fold more
NO2 in the circuit from the source cylinder than
is formed spontaneously in the circuit. On the basis of this analysis,
it is important that NO source mixture manufacturers focus on
production practices that minimize the amount of
NO2 in the source cylinders.
It is possible to accidentally contaminate NO source cylinders with NO2 formed via a back diffusion mechanism when the cylinder valve is opened to an improperly purged regulator containing trace amounts of room air. Vacuum purging of regulators to remove ambient air before opening the cylinder valve can minimize any potential for accidental contamination of the cylinder. The importance of accidental contamination is difficult to evaluate. However, until data are available, the appropriate purging and handling of NO cylinders is recommended.
Whenever NO is in the presence of oxygen, additional NO2 production can occur. The dwell time within the conductive airways will add to the amount of NO2 delivered to the patient. Additional areas of concern yet to be investigated include the production of NO2 at the alveolus.
In summary, NO2 formation is minimal during inhaled NO
therapy in the continuous flow neonatal ventilatory circuit described
in Fig. 1
, circuit A. Existing chemical kinetics and rate constants can
be used to reliably predict the amount of NO2 formed in a
continuous flow system. At fixed NO and oxygen concentrations, the
amount of NO2 formed will be dependent upon the dwell time.
Investigators using inhaled NO therapy should calculate dwell times and
the potential NO2 delivery to the patient. Additional
sources of NO2 (NO source cylinder, cylinder contamination,
NO2 formation within the conductive airways and at the
alveolus) need to be considered when estimating the total
NO2 delivery to the patient.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  K. P. Van Meurs Inhaled Nitric Oxide Therapy in the Preterm Infant Who Has Respiratory Distress Syndrome NeoReviews, June 1, 2005; 6(6): e268 - e277. [Full Text] [PDF]
|
|
|
|
 |
|
 F. Ichinose, J. D. Roberts Jr, and W. M. Zapol Inhaled Nitric Oxide: A Selective Pulmonary Vasodilator: Current Uses and Therapeutic Potential Circulation, June 29, 2004; 109(25): 3106 - 3111. [Full Text] [PDF]
|
|
|
|
 |
|
 T. KOBAYASHI, E. C. GABAZZA, S. SHIMIZU, H. YASUI, H. YUDA, O. HATAJI, K. MARUYAMA, T. YAMAUCHI, K. SUZUKI, Y. ADACHI, et al. Long-term Inhalation of High-dose Nitric Oxide Increases Intraalveolar Activation of Coagulation System in Mice Am. J. Respir. Crit. Care Med., June 1, 2001; 163(7): 1676 - 1682. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Weinberger, D. L. Laskin, D. E. Heck, and J. D. Laskin The Toxicology of Inhaled Nitric Oxide Toxicol. Sci., January 1, 2001; 59(1): 5 - 16. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
