Subscriber access provided by Queen Mary, University of London
Article
Measured saturation vapour pressures of phenolic and nitro-aromatic compounds Thomas J Bannan, Alastair Murray Booth, Benjamin T Jones, Simon O'Meara, Mark Howard Barley, Ilona Riipinen, Carl John Percival, and David O. Topping Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b06364 • Publication Date (Web): 06 Mar 2017 Downloaded from http://pubs.acs.org on March 7, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 19
Environmental Science & Technology
1
Measured saturation vapour pressures of
2
phenolic and nitro-aromatic compounds
3
1*
Thomas J Bannan, 1A Murray Booth, 1Benjamin T Jones, 1Simon O’Meara, 1Mark H Barley, 3Ilona
4
Riipinen, 1Carl J. Percival, 1,2David Topping.
5
1
6
UK
7
2
National Centre for Atmospheric Science, University of Manchester, Manchester, UK.
8
3
Department of Environmental Science and Analytical Chemistry (ACES), Stockholm University,
9
Stockholm, Sweden.
School of Earth, Environmental and Atmospheric Science, University of Manchester, Manchester,
10
*Corresponding author: Thomas J. Bannan. 4.30 Simon Building, University of Manchester, Oxford
11
Road, Manchester, M13 9PL, Phone: 0161 306 6587, Fax: 0161 306 3951, Email:
12
[email protected] 13
TOC Art
14 15
Abstract
16
Phenolic and nitro-aromatic compounds are extremely toxic components of atmospheric aerosol
17
that are currently not well understood. In this paper, solid and sub-cooled liquid state saturation
18
vapour pressures of phenolic and nitro-aromatic compounds are measured using Knudsen Effusion
19
Mass Spectrometry (KEMS) over a range of temperatures (298–318 K). Vapour pressure estimation
20
methods, assessed in this study, do not replicate the observed dependency on the relative positions
21
of functional groups. With a few exceptions, the estimates are biased towards predicting saturation
22
vapour pressures that are too high, by 5 to 6 orders of magnitude in some cases. Basic partitioning
ACS Paragon Plus Environment
Environmental Science & Technology
23
theory comparisons indicate that overestimation of vapour pressures in such cases would cause us
24
to expect these compounds to be present in the gas state, whereas measurements in this study
25
suggest these phenolic and nitro-aromatic will partition into the condensed state for a wide range of
26
ambient conditions if absorptive partitioning plays a dominant role. Whilst these techniques might
27
have both structural and parametric uncertainties, the new data presented here should support
28
studies trying to ascertain the role of nitrogen containing organics on aerosol growth and human
29
health impacts.
30
Introduction
31
Aerosol particles are an important component of the Earth’s atmosphere, affecting visibility, human
32
health and climate on regional and global scales (Bilde et al., 2015). Mechanistic models attempt to
33
predict the evolving composition and microphysics of aerosol populations through the use of
34
fundamental properties of pure components and mixtures. A critical parameter in this regard is the
35
pure component equilibrium vapour pressure (here referred to as saturation vapour pressure) of the
36
constituent molecules. Given the chemical complexity of the organic fraction of atmospheric aerosol,
37
with potentially many hundreds of thousands of individual species (Jenkin et al., 2003, Aumont et al.,
38
2005) and the scarcity of existing experimental data, the saturation vapour pressure of a large
39
fraction of atmospherically relevant compounds must be estimated rather than comprehensively
40
measured. There are a number of estimation methods available (e.g. Joback et al., 1987; Stein and
41
Brown 1994; Myrdal and Yalkowsky, 1997), but these suffer from inherent bias due to the lack of
42
data for heavily oxygenated, low volatility multifunctional compounds in the fitting dataset, which
43
are the sort of molecules expected to be present in atmospheric aerosols (Bilde et al., 2015). This
44
has been recognised as a problem in predicting total mass and speciation (Barley and McFiggans,
45
2010, Booth et al., 2010, O’Meara et al., 2014) and there have been efforts to develop improved
46
estimation methods more suited to atmospherically relevant compounds (e.g. EVAPORATION
47
(Compernolle et al., 2011)). However, such methods still remain largely invalidated across broad
48
ranges in absolute saturation vapour pressure and functionality. The most recent comparison
49
between methods was carried out by O’Meara et al (2014), using a database of 90 compounds
50
covering a vapour pressure range of 60 Pa to 1.34x10-5 Pa at 298 K. Whilst a first order problem
51
might be the propagated uncertainty in total particulate mass, it is important to evaluate any
52
heterogeneity in predicting the concentration of certain chemical classes. Developing accurate
53
predictive methods is challenging due to the influence of additional groups not necessarily being
54
linear (intramolecular bonding, shielding or steric hindrance), and the spread between data from
55
different instruments (Bilde et al., 2015).
ACS Paragon Plus Environment
Page 2 of 19
Page 3 of 19
Environmental Science & Technology
56
With this in mind, this paper adds to the existing body of work in which we measure the saturated
57
vapour pressure of two related classes: naphthalene phenolic and nitro-aromatic compounds.
58
Oxygenated aromatics (acids and phenols) have been measured in atmospheric aerosols (Hemming
59
and Seinfeld 2001; Kawamura and Kaplan 1987; Yassineet al., 2009). Phenolic compounds are useful
60
tracers for anthropogenic emissions as they are produced from a variety of anthropogenic volatile
61
organic compound (VOC) precursors such as benzene, toluene and xylene (Tremp et al., 1993;
62
Berndt and Böge, 2006) and a major component of biomass burning (Yee et al., 2013). For resolving
63
discrepancies between instruments and models they are also useful for probing intramolecular
64
effects due to the relative ease of obtaining samples with phenol groups in different configurations.
65
Nitro-aromatic compounds have proved to be useful tracers for anthropogenic emissions (Grosjean,
66
1992) and biomass burning (Yee et al., 2013), whilst studies quantifying the overall role of nitrogen-
67
containing organic compounds on aerosol growth could benefit from more refined pure component
68
vapour pressures (Smith et al 2008; Duporte et al 2016). With regards to potential impacts, many of
69
these compounds are noted to be highly toxic (Kovacic and Somanathan, 2014). Relating
70
toxicological response to certain classes is not the focus of this study, yet growing evidence suggests
71
specific particle components might be more important for certain diseases (Cassee et al., 2013,
72
Atkinson et al., 2010) whilst the majority of correlating studies rely on existing epidemiological
73
techniques (Stanek et al., 2011). Even if mechanistic models struggle to predict adequate mass due
74
to missing process phenomena (e.g.McVay et al., 2016), resolving expected partitioning in measured
75
samples is important. More generally the data presented should lead to a more thorough evaluation,
76
and potentially improvement, of vapour pressure predictive techniques as there is a significant lack
77
of data for functionalised nitrogen containing organic compounds in the datasets used for fitting
78
model parameters (e.g. Nannoolal et al., 2008).
79
In the following text we first briefly present the experimental and modelling methodology before
80
discussing measured vapour pressure from a Knudsen Effusion Mass Spectrometer (KEMS) (Booth et
81
al., 2012) and comparisons with a range of estimation methods. We then discuss the need for
82
measurements from complimentary techniques to improve datasets used for refitting estimation
83
methods.
84
2.Experimental
85
The following compounds were purchased from Sigma-Aldrich and used with no further preparation:
86
1-Naphthol,
87
dihydroxynaphthalene, 2,7-dihydroxynaphthalene, 1,5-dihydroxynaphthalene, 1-nitronaphthalene,
88
1,3-dinitronaphthalene, 2-methyl-1-nitronaphthalene, ortho-amino-benzoic acid, meta-amino-
2-
Naphthol,
1,3-dihydroxynaphthalene,
2,3-dihydroxynaphthalene,
ACS Paragon Plus Environment
1,7-
Environmental Science & Technology
Page 4 of 19
89
benzoic acid, para-amino-benzoic acid, meta-nitrophenol, para-nitrophenol, ortho-nitroaniline,
90
meta-nitroaniline, para-nitroaniline and Glutarimide, all of which are solid at room temperature.
91
2.1 The Knudsen Effusion Mass Spectrometry (KEMS)
92
The KEMS system is identical to that used in previous studies (e.g. Booth et al., 2009) and an
93
overview of the measurement procedure is repeated here: to calibrate, a sample of known vapour
94
pressure is placed in the temperature controlled Knudsen cell, in this case para-anisic acid (VP at
95
298k = 5.24x10-4 Pa, (Colomina et al., 1978; Booth et al., 2012). The linearity of the one calibration
96
point for calibrating measurements over a wide range of vapour pressures has been tested
97
exhaustively for measurements at close to limits of detection of the instrument (1 × 10-8 Pa) and for
98
higher vapour measurements such as benzophenone. This was most recently validated through
99
comparison with other instruments, in work that will be presented at a later date, which presents
100
measurements agreeing well across techniques over a vapour pressure range of 1x10-2-1x10-7 (Pa).
101
The cell has a chamfered effusing orifice with a size ≤1/10 the mean free path of the gas molecules
102
in the cell. This ensures the orifice does not significantly disturb the thermodynamic equilibrium of
103
the samples in the cell (Hilpert, 2001). The resulting molecular beam is ionised by 70 eV electron
104
impact, then sampled by the mass spectrometer. After correcting for the ionization cross section of
105
the calibration compound, this produces a signal proportional to the vapour pressure.
106
After this calibration a sample of unknown vapour pressure can be measured. During sample
107
change, the first chamber with the Knudsen cell is isolated via the gate valve and vented to air
108
allowing the ioniser filament and secondary electron multiplier to be left on. The unknown vapour
109
pressures can be determined from the intensity of the mass spectrometer signal of the compound in
110
question. If the Knudsen number, the ratio of the mean free path of molecules to the size of the
111
effusion orifice, is high enough, then effusing gas does not significantly disturb the equilibrium in the
112
cell (Booth et al., 2009, Hilpert 1991; 2001). This allows the steady state vapour pressure, as
113
measured by the KEMS, to be as close as possible to the equilibrium vapour pressure.
114
Once the solid state vapour pressure, P (Pa), has been determined at a number of different
115
temperatures, the Clausius-Clapeyeron equation is used to derive enthalpies and entropies of
116
sublimation as described in Booth et al., (2009);
117
ln ܲ =
௱ுೞೠ್ ோ்
+
௱ௌೞೠ್ ோ
(1)
118
where T is the temperature (K), R is the ideal gas constant (J mol−1 K−1) and ΔHsub (J mol-1) and ΔSsub (J
119
mol-1 K-1) are the enthalpies and entropies of sublimation respectively. P (Pa) was obtained over a
ACS Paragon Plus Environment
Page 5 of 19
Environmental Science & Technology
120
range of 20 K in this work, starting at 298 K. The reported solid state vapour pressures at 298.5K
121
(P298) are calculated from the linear fit of ln P vs. 1/T used in the Clausius-Clapeyon equation.
122
2.2 Differential Scanning Calorimetry (DSC)
123
As noted in the proceeding section, since we measure the saturation vapour pressure above the
124
solid state, we need to convert this to sub-cooled reference state for comparison with predictive
125
techniques and use within dynamic aerosol models. Following previous studies (e.g. Booth et al.,
126
2010), the information needed to perform this conversion, melting points (Tm) and Enthalpies of
127
Fusion (ΔHfus), were measured using a TA instruments Q200 Differential Scanning Calorimeter (DSC).
128
The procedure is repeated here: heat flow and temperature were calibrated using an indium
129
reference, and heat capacity using sapphire reference. A heating rate of 10 K min-1 was used. 5 – 10
130
mg of sample was measured using a microbalance, then pressed into a hermetically sealed
131
aluminium DSC pan. A purge gas of N2 was used with a flow rate of 30 ml min-1. The reference was
132
an empty sealed pan of the same type. Data processing was performed using the `Universal Analysis'
133
software supplied with the instrument. Δcp,sl was estimated using Δcp,sl = ΔSfus (Mauger et al., 1972;
134
Grant et al., 1984).
135
3. Theory
136
3.1 Sub-cooled correction
137
The sub-cooled vapour pressure is derived from the value measured above the solid state using the
138
Prausnitz equation (Prausnitz et al., 1986):
139
ln
Pl ∆H fus Tm ∆c p , sl Tm ∆c p , sl Tm = ln − 1 − − 1 + Ps RTm T R T R T
(2)
140
where: P is the saturation vapour pressure (Pa) with the subscript ‘s’ referring to the solid and ‘l’ to
141
the sub-cooled liquid phase, ΔHfus is the enthalpy of fusion (J mol-1), Δcp,sl denotes the best estimate
142
of the underlying change in heat capacity between the liquid and solid state at the melting point (J
143
mol-1 K-1), T is the temperature (K), Tm is the melting point (K) (which is commonly used instead of
144
the triple point,(Tt) (Prausnitz et al., 1986).
145
3.2 Vapour pressure predictive techniques
146
The predictive techniques used here are the same as those in a recent evaluation paper by O’Meara
147
et al., (2014) with the exception of the EVAPORATION method which is currently only applicable to
148
aliphatic compounds. These methods are all based on group contributions where each group has a
ACS Paragon Plus Environment
Environmental Science & Technology
149
set of values associated with it for calculating the appropriate property obtained by fitting to data.
150
The accuracy of a given method is therefore very sensitive to the choice of data used to fit it. This is
151
especially true for groups where data is limited, for which this study is relevant.
152
Nine methods are evaluated in total. Four of the group contribution vapour pressure methods
153
extrapolate from a boiling point down to the temperature of interest. These four methods are;
154
Nannoolal et al (2008), Moller et al (2008), Myrdal and Yalkowsky (1997) and Lee-Kesler (Reid et al.,
155
1987). The two boiling point estimation methods that are then used are Nannoolal et al., (2004) and
156
Stein and Brown (1994), for a total of 8 methods. The method of Joback (1987) is excluded due to its
157
known biases (Barley and McFiggans, 2010). The ninth method is SIMPOL (Pankow and Asher, 2008)
158
that directly estimates the vapour pressure without requiring a boiling point value.
159
4. Results
160
4.1 Solid state vapour pressures
161
Solid state vapour pressures directly measured in the KEMS are reported in Table 1. Measurements
162
were made from 298 to 318 +/- 274 K and the Clausius-Clayperon equation was used to derive
163
enthalpies and entropies of sublimation, as noted in section 1. 1,3 and 2,3 dihydroxnapthalene
164
were measured up to 333 K to produce enough vapour, thus signal, to detect with the KEMS, and the
165
P298 was extrapolated from these measurements. However, 1,7/2,7/1,5-dihydroxynaphthalene could
166
not be detected even at elevated temperatures, from which we infer that the vapour pressure is less
167
than the current reported minimum detectable pressure in the KEMS which corresponds to 1 × 10-8
168
Pa as determined by Booth et al., (2009). Some general observations regarding functional group
169
positioning on the solid-state vapour pressure can be inferred from table 1. 1,3-
170
dihydroxynaphthalene is within the current reported detectable limit (P298 of 9 × 10-6 Pa) whilst 2,3-
171
dihydroxynaphthalene has a volatility two orders of magnitude higher than the compound with the
172
functional group at the 1,3 position. One possible explanation for this is intramolecular bonding,
173
where adjacent groups on a molecule can hydrogen bond with each other thus lowering the
174
apparent polarity of the molecule and raising the apparent volatility relative to a molecule with non-
175
adjacent groups. There is no discernable pattern for the positional effects (ortho-meta-para
176
isomerism) in the amino-benzoic acids, nitrophenols and nitroanilines. Previous work on different
177
ortho-meta-para substitutions of benzoic acid showed that in molecules with different functional
178
groups the isomer effects were the result of a complex interplay of steric, resonance effect and
179
polarity (Booth et al., 2012). As noted in Bilde et al (2015), measurements between different
ACS Paragon Plus Environment
Page 6 of 19
Page 7 of 19
Environmental Science & Technology
180
instruments can often diverge; resolving the true effect of functional group positioning is still not
181
clear and might benefit from quantum scale modelling (Schroeder et al., 2016, Kurten et al., 2016).
182 183
ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 19
184
Table 1. Solid state vapour pressure at 298 K, enthalpies and entropies of sublimation. Estimated
185
errors on the vapour pressures are
