Heard D.E. (editor) Analytical Techniques for Atmospheric Measurement

Подождите немного. Документ загружается.

418 Analytical Techniques for Atmospheric Measurement

intervals  over the wavelength range that contributes to the photolysis process of

interest. Photolysis frequencies are then calculated from the measured radiation spectra



 and the relevant molecular cross section  and quantum yield , according to

Equation 9.12. The main advantage of spectroradiometry is the possibility to determine

photolysis frequencies for a whole set of different molecules at the same time.

•

Actinic-flux filter radiometers have a broad spectral coverage and measure the actinic

flux integrated over all wavelengths simultaneously. The optical wavelength selection is

chosen to mimic the spectral function  × of the relevant photolysis frequency by

means of an optical transmission filter and a photoelectric sensor. In a first approxi-

mation, the filter radiometer signal is proportional to the photolysis frequency and can

be calibrated against an absolute reference (e.g. a chemical actinometer). In order to

obtain more accurate results, the dependence of the calibration on ambient conditions

(e.g. solar zenith angle, temperature) must be taken into account. Filter radiometers

are generally useful because of their compact size and fast time response.

•

Irradiance radiometers (spectroradiometers and broadband radiometers) can be used for

the determination of photolysis frequencies in principle like actinic-flux instruments.

However, the measured irradiances must be converted into actinic flux data, a step

that introduces additional errors in the derived photolysis frequencies. The conversion

requires knowledge of the angular distribution of the radiance in the atmosphere.

The latter information is difficult to obtain in the field, especially in the presence

of clouds and aerosols, and is usually estimated by empirical rules or by numerical

models. Irradiance radiometers have the advantage that they have a wide distribution

as meteorological instruments and are readily available as standard equipment from

various manufacturers. For details, see Section 9.6.

9.2.3 Actinic-ﬂux sensing techniques

For practical purposes, most important are the measurement techniques that are directly

sensitive to the actinic flux, that is have the ability to collect radiation with equal

sensitivity independent of the angle of incidence (isotropic behaviour). To this class

of techniques belong the abovementioned chemical actinometers, actinic-flux spectrora-

diometers and actinic-flux filter radiometers. The measurement of actinic flux requires

the integral detection of the incident radiation from all directions over a solid angle of

4 sr. Spherical or tubular photolysis reactors used in chemical actinometry closely fulfil

this requirement (Figure 9.8A). Actinic flux radiometers mostly use hemispherical input

optics, such that two instruments must be combined to cover 4 sr (Figure 9.8B). The

setup of two radiometers pointing into opposite directions has the additional advantage

of providing separate information about the up- and down-welling components of the

actinic radiation.

The basic features of actinic-flux sensing techniques are summarized in Table 9.2, while

detailed descriptions follow in later sections. It may be noted that chemical actinometry

is the oldest actinic-flux method and has been used since the mid-1970s to measure

photolysis frequencies in the atmosphere, mainly for NO

and O

(see Section 9.3).

Actinic-flux radiometry was developed later, starting at the end of the 1980s, and is now

widely applied in atmospheric chemistry for measurement of photolysis frequencies of

Measurement of Photolysis Frequencies in the Atmosphere 419

Gas inlet

(a) (b)

Quartz bulb

Artificial horizon

Photoelectric

detector

unit

Opt.

receiver

Figure 9.8 Receiver geometries for solar actinic ﬂux. (a) Photolysis reactor of a chemical actinometer

with a spherical response. (b) Combination of two hemispheric radiometers facing into opposite directions.

various gases (see Sections 9.4, 9.5, 9.8). The performance and reliability of actinic-flux

techniques have been thoroughly tested in several assessments. Examples of results are

given in Section 9.8.3.

9.2.4 Radiative transfer models

For completeness, theoretical modelling should be mentioned as a method to determine

photolysis frequencies.

Radiative transfer (RT) models calculate first and foremost the spectral actinic flux in

the atmosphere. The models use the extraterrestrial solar spectrum as input, as well as

optical properties of the atmosphere and the earth’s surface albedo. In a second step,

photolysis frequencies can be calculated from the simulated radiation spectra F



 and

the relevant molecular data  and , using Equation 9.12.

Radiative transfer models play an important role for the prediction of photolysis

frequencies in atmospheric chemistry. In particular, they provide a broad temporal

and spatial coverage that currently cannot be achieved by measurements alone. On

the other hand models have difficulties to calculate accurate j-values at local sites

where the field-of-view is shadowed by trees, buildings or mountains, or when complex

cloudiness prevails. Model results are therefore no good substitute for in situ measure-

ments in local photochemistry field campaigns. However, models are useful to assist in

420 Analytical Techniques for Atmospheric Measurement

Table 9.2 In situ measurement techniques for atmospheric photolysis frequencies

Feature Chemical

actinometer

Actinic ﬂux

spectroradiometer

Actinic ﬂux ﬁlter

radiometer

Main components Spherical or

cylindrical quartz

reactor + gas

analysis

Optical receiver +

scanning

monochromator or

spectrograph +

photoelectric

detection system

Optical receiver +

optical ﬁlter +

photoelectric detector

Measured quantity Concentration [AB] F



 U =





d

Signal evaluation j =−

t

AB

AB

j =



F





j = AU

Versatility Single speciﬁc

photolysis reaction

Various photolysis

reactions at the

same time

Single speciﬁc

photolysis reaction

Reported applications O

,NO

, RONO

 NO

 HONO

HCHO

,NO

Calibration reference Gas standard Calibration lamp Reference instrument

or calibration lamp

Sensitivity Medium to good Very good Good

Time resolution Flow reactors:

1–60 s; Static

reactors: 1–60 min

Scanning

monochromator

systems: 15–120 s;

Spectrograph

systems: 1–10 s

∼ 1s

Size Large to medium;

requires

pressurised gases

Medium to small Small

Listed are techniques that collect radiation with isotropic sensitivity.

U is a voltage signal; D

represents the wavelength-dependent sensitivity of the ﬁlter radiometer; the spectral design

of the ﬁlter radiometer aims for D

 ∝  and requires knowledge of  and .

Equation 9.1, or alternatively equation 9.3 may be applied.

Equation 9.12 is applied; it requires knowledge of  T  and  T , and, if necessary, of ambient temperature T .

See equation 9.45; the calibration factor A usually depends on ambient parameters such as solar zenith angle,

temperature etc.

Reference instrument can be a chemical actinometer, a spectroradiometer, or another calibrated ﬁlter radiometer.

the determination of photolysis frequencies by certain techniques, like filter radiometry

or irradiance radiometry. See Section 9.7 for more information about RT modelling.

9.3 Chemical actinometry

Chemical actinometry in its original sense means the quantification of an absolute photon

flux by measuring the rate of photoinduced change in a chemical system (Calvert & Pitts,

1966; Pravilov, 1987; Kuhn et al., 2004). In general the chemical system can be a solid,

Measurement of Photolysis Frequencies in the Atmosphere 421

liquid or gas and is called chemical actinometer. Actinometry is then based on the principle

that the photolytical conversion rate of molecules in an actinometer cell is equal to the

absorption rate of photons times the quantum yield of the actinometer. The absorption

rate of photons in the actinometer is related to the incident photon flux by Lambert–

Beer’s law. Application of these fundamental principles allows the calculation of absolute

incident photon-fluxes from measured chemical-conversion rates, if the quantum yield

of the actinometer is well known (for details, see Calvert & Pitts, 1966).

While classical chemical actinometry has been used in laboratories by photochemists

for almost a century (e.g. Warburg, 1912; Vaughan & Noyes, Jr., 1930), the general

concept has been adapted, beginning in the 1970s, by atmospheric chemists for the

measurement of photolysis frequencies in the atmosphere. Applications have been

reported for O

,NO

and alkyl nitrates (Tables 9.3 and 9.4). Corresponding techniques for

the photolysis of O

and NO

will be presented in some detail in Sections 9.3.4 and 9.3.5,

respectively.

9.3.1 Principle of chemical actinometry

Chemical actinometers for atmospheric measurements are realised by (1) filling the gas

of interest into a photolysis reactor, (2) exposing the actinometer to solar radiation,

and (3) measuring the photochemical conversion rate. Here, it is important that the

actinometer gas is exposed to the ambient actinic flux without altering the spectral

composition and intensity of the radiation significantly. This is achieved by the usage

of a transparent quartz cell with a suitable geometrical shape (spherical or cylindrical)

and by applying gas concentrations with small optical absorbances (see more about this

aspect in Section 9.3.3 and Appendix A.2). If these conditions are fulfilled, the photolysis

frequency can be evaluated according to Equation 9.1, which requires measurements of

gas concentrations and time but which does not need the knowledge of the absorption

spectrum and quantum yield of the primary photodissociation.

There are two possible modes of actinometer operation. In the static batch mode the

actinometer gas is filled into the photolysis reactor which is then sealed off by a gas

valve and is covered with an opaque hood to exclude sunlight. For measurement, the

actinometer is uncovered and exposed to solar radiation for a fixed time interval t.

Thereafter, the actinometer is covered again and is analysed for change in its chemical

composition.

The other mode of operation involves flowing gas. The actinometer gas is passed

continuously through a photolysis reactor which is permanently exposed to ambient solar

radiation. In this case t is equal to the mean residence time of the gas in the illuminated

photolysis cell. After having passed the reactor, the gas composition is analysed by an

online gas detector. This operational mode allows a continuous monitoring of photolysis

frequencies.

9.3.2 Chemical kinetics considerations

The rate of photochemical change for a general reaction AB +h → A +B can be

determined from either the loss of reactant AB or the increase of its product A (or B).

Table 9.3 O

→ O

D chemical actinometers used in atmospheric ﬁeld studies

Reference Photolysis reactor

Gas mixtures

Exposure time Detection technique k

Platform

Bahe and

Schurath (1978);

Bahe et al.

(1979)

Static and ﬂow

types, spher.

O 0.5–2 h N

: gas chromatography 062 Ground + balloon

h =26 km

Dickerson et al.

(1979, 1982)

Flow type, cyl. O

O10s N

: conversion into

NO + detection by

chemiluminescence

08 Ground + aircraft

h =01–68km

Junkermann

et al. (1989)

Static type, spher. O

O <30 min N

: gas chromatography 075 Ground

Blackburn et al.

(1992)

Flow type, cyl. O

O34s NO

: transfer into

liquid methanol + detection

by electrical conductivity

079 Ground

Bairai and

Stedman (1992)

Flow type, cyl. O

O/He 7 s NO

: transfer into luminol

solution + ﬂuorescence

detection

08 Ground

Müller et al.

(1995)

Static type, spher. O

O 0.8–2 h N

: gas chromatography 075 Ground

Shetter et al.

(1996)

Flow type, cyl. O

O40s N

: transfer into liquid

water + detection by

electrical conductivity

072 Ground

Cyl., cylindrical tube; spher., spherical bulb. Typical reactor volumes are 50–250cm

. Walls are made of transparent quartz.

N

OO

>100; total pressure is about 1 atm in ﬂow systems, and 1–2 atm in static systems.

Branching ratio of the O

D +N

O reaction channels producing (a) N

and (b) NO+NO, used for the calculation of the jO

D-values.

Table 9.4 NO

→ NO +O and RONO

→ RO +NO

chemical actinometers used in atmospheric ﬁeld studies

Reference Photolysis reactor

Gas mixtures

Exposure time Detection technique

Platform

→ NO +O actinometer

Jackson et al.

(1975); Harvey

et al. (1977)

Flow type, cyl. NO

/Air 1–4 s NO, NO

chemiluminescence

Ground

Zafonte et al.

(1977)

Flow type, cyl. NO

7 s NO, NO

chemiluminescence

Ground

Bahe et al. (1980) Flow type, spher. NO

10 s NO: chemiluminescence Ground

Static type, cyl. NO

30 s NO

: spectrophotometry Ground

Dickerson et al.

(1982)

Flow type, cyl. NO

1 s NO, NO

chemiluminescence

Ground + aircraft

h =01–55km

Parrish et al. (1983) Flow type, cyl. NO

(60 hPa, total)

5 s NO, NO

chemiluminescence

Ground

Madronich et al.

(1983, 1985)

Static type, cyl. NO

5–10 min Pressure increase Ground + balloon

(h =24, 32 km)

Shetter et al.

(1992); Lantz et al.

(1996)

Flow type, cyl. NO

(67 hPa, total)

0.3–0.5 s NO, NO

chemiluminescence

Ground

Schultz et al.

(1995)

Static type, cyl. NO

1–2 min NO

: photometry Ground

Table 9.4 (Continued)

Reference Photolysis reactor

Gas mixtures

Exposure time Detection technique

Platform

Kelley et al. (1995);

Dickerson et al.

(1997)

Flow type, cyl. (dual

system, upward and

downward looking)

/Air 1 s NO, NO

chemiluminescence

Aircraft

h =02–76km

Castro et al. (1995,

1997)

Flow type, cyl. NO

1 s NO, NO

chemiluminescence

Ground

Vuilleumier et al.

(2001)

Flow type, cyl. NO

1 min NO, NO

chemiluminescence

Ground

Kraus et al. (2000) Flow type, cyl. NO

/Air 1 min NO, NO

chemiluminescence

Ground

RONO

→ RO +NO

actinometer

Luke et al. (1989) Flow type, cyl. RONO

/Air

6–11 s NO

: transfer into luminol

solution + ﬂuorescence

detection

Ground

Cyl., cylindrical tube; spher., spherical bulb. Typical reactor volumes are 7 to 380 cm

. Walls are made from transparent quartz.

In NO

ﬂow actinometers the mixing ratio of NO

is typically between 1 and 80 ppmv at total pressures around 1 atm, unless otherwise indicated. In static systems pure NO

is used

at pressures of 1–5 hPa.

can be detected by chemiluminescence after chemical conversion into NO.

R = ethyl, n-propyl, n-butyl, 2-butyl; mixing ratios of RONO

are between 23–52 ppmv.

Measurement of Photolysis Frequencies in the Atmosphere 425

If the exposure time t is held at less than 1% of the reciprocal j-value, the concentration

of AB will change very little and AB can be assumed to have a constant concentration

AB ≈ AB

. It is then reasonable to measure the increase of the product A, which can

be related to the photolysis frequency as follows:

dA =jABdt (9.19)

Integration over the time interval t yields

A = AB



t

jt d t (9.20)

In principle the photolysis frequency may vary during the exposure time t. It is then

useful to define a mean value

j for this time interval:

j =

t



t

jt d t (9.21)

Combining Equations 9.20 and 9.21 results in

j =

AB

A

t

(9.22)

This demonstrates that the photolysis frequency derived from the measured increase A

is a true mean value over the respective time interval.

Another approach is to let the initial concentration AB

noticeably decay and measure

the decrease AB during t. In this case we get

dAB =−jAB dt (9.23)

which after integration yields

j =−

t

AB

−AB

AB

(9.24)

Here again, a mean j-value is obtained for the time of exposure.

It should be noted that the photochemistry and its kinetics in real chemical actinometers

is often complicated by secondary chemistry. The reason is that the photolytical products,

atoms and radicals, are highly reactive. Consecutive reactions in the gase phase and on

wall surfaces lead to losses of the products, and possibly also of the initial reactant. For an

accurate evaluation of the required photolysis frequency the effect of secondary chemistry

must be taken into account. Sometimes scavenger gases are added to the actinometer gas

in order to capture the reactive primary photofragments and convert them into species

which are more stable or easier to measure. Examples will be given in Sections 9.3.4

and 9.3.5.

426 Analytical Techniques for Atmospheric Measurement

9.3.3 Optical and photochemical considerations

The advantage of a properly designed chemical actinometer is that the gas molecules

inside the photolysis reactor will photodissociate with the same rate coefficient as would

molecules do in the surrounding atmosphere. This will only happen if the conditions that

have an influence on the photolysis rate are the same inside and outside of the chemical

actinometer.

Gas temperature and pressure have a possible influence on the absorption cross section

and quantum yield of photolysis reactions. For example, the photodissociation of O

→

O

D exhibits a strong temperature dependence (Dickerson et al., 1982; Bohn et al.,

2004), while the NO

photolysis is only weakly temperature dependent (Dickerson et al.,

1982; Shetter et al., 1988). In few cases, a notable pressure dependence has been observed,

such as in the formation of H

+CO in the photolysis of HCHO (Moortgat et al., 1983).

Whenever a temperature or pressure dependence exists, the operating conditions in the

actinometer must be chosen similar to the corresponding ambient conditions in order to

provide representative measurements.

Optical effects caused by the design of the chemical actinometer can modify the actinic

flux inside the photolysis reactor. Potentially-interfering processes are:

•

Strong absorption by the actinometer gas.

•

Absorption by the photolysis-cell wall-material.

•

Optical reflections at the outer and inner wall surfaces.

•

Shielding of solar radiation by the gas inlet/outlet of the photolysis cell.

Gas phase absorption can be a problem at high concentrations of the actinometer

gas. If the gas phase is optically thick, for example the absorbance exceeds 1%, an

appreciable fraction of radiation is absorbed in the photolysis cell. Correspondingly,

the absorption protects some of the actinometer gas from photolysis and the resulting

photolysis frequency will be systematically lower than in the surrounding atmosphere.

Optically thin conditions can be achieved by a sufficiently low actinometer-gas concen-

tration, which in turn may require the use of highly sensitive analytical instruments for

the measurement of correspondingly low reactant and product concentrations.

Chemical-actinometer walls are usually made of quartz and have a typical thickness

of a few millimeters. At this thickness the material is highly transparent and has an

absorbance of less than 1% between 250 and 1000 nm. Thus, losses of light due to wall

absorption are negligible. A considerable amount of incident solar radiation is, however,

lost by reflection I

 when the light enters the photolysis reactor (Figure 9.9; for details,

see Appendix A.2). The reflectance of the quartz wall depends on the angle of incidence

 and varies for unpolarized UV radiation between ∼8% at normal incidence  = 0



to 100% at grazing incidence  = 90



. The attenuated radiation I

= I

−I

 entering

the photolysis cell is subsequently reflected multiple times at the inner surface of the

photolysis cell. These reflections enhance the photon flux inside the chemical actinometer.

For a photolysis reactor that has a spherical or infinitely long cylindrical shape and

non-absorbing walls, it has been shown in a theoretical study by Zafonte et al. (1977)

that the reflections at the exterior and interior wall surfaces cancel each other exactly,

Measurement of Photolysis Frequencies in the Atmosphere 427

Figure 9.9 Illustration of multiple reﬂections inside a spherical or cylindrical (inﬁnite length) photolysis

reactor. I

is the intensity of the incident solar radiation. I

is the intensity lost by reﬂection at the outer

wall and I

is the transmitted intensity. I

I

 denote the intensities of the internally reﬂected light.

that is I

+···=I

(see Appendix A.2). Therefore, the actinic flux inside the

actinometer is expected to be the same as in the surrounding atmosphere. The perfect

geometry of the photolysis cell is, however, perturbed by the gas inlet/outlet (usually the

ends, in case of cylindrical tubes) which act as a sink for reflected radiation and shield

some of the incident solar radiation. It is generally assumed that the corresponding loss

of light is small and equal to the ratio of the area of the inlet/outlet to the total surface

area of the photolysis reactor (Zafonte et al., 1977).

The theoretical prediction of the cancelling effects of external and internal reflections

has been confirmed experimentally (Dickerson & Stedman, 1980; Madronich et al., 1983).

In one study, one to three additional cylindrical tubes were slid over the initial photolysis

tube in which the photolysis frequency of NO

was measured in a field experiment. Within

5% precision of the experiment, the additional tubes had no effect on the measured

j-value, suggesting that a single tube caused less than 1.7% loss of radiation inside the

actinometer (Dickerson & Stedman, 1980).

9.3.4 Ozone chemical actinometers

The photodissociation of O

in the UV and visible region has five energetically possible

pathways for which experimental evidence exists (Matsumi & Kawasaki, 2003):

+h < 310 nm −→ O

D +O

a



 (R 9.4)

+h < 411 nm −→ O

D +O

X



−

 (R 9.5)

+h < 463 nm −→ O

P +O

b



 (R 9.8)

+h < 612 nm −→ O

P +O

a



 (R 9.9)

+h < 1180 nm −→ O

P +O

X



−

 (R 9.10)