Raman spectroscopy and atomic force microscopy of the reaction of sulfuric acid with sodium chloride

Article abstract of DOI:10.1021/ja001469x

The reaction of NaCl with H₂SO₄ is investigated using Raman spectroscopy, atomic force microscopy (AFM), and gravimetric analysis. Raman spectra are consistent with the formation of NaHSO₄ with no evidence for Na₂SO₄. The spectra indicate that the phase of NaHSO₄ varies with the amount of H₂O in the H₂SO₄. At low H₂O concentrations, the reaction produces anhydrous β-NaHSO₄, which undergoes a phase change to anhydrous α-NaHSO₄ over the course of 18 h. At higher H₂O concentrations, anhydrous α-NaHSO₄ is formed with small amounts of NaHSO₄·H₂O. AFM measurements on NaCl (100) show the formation of two distinct types of NaHSO₄ structures consistent in shape with α-NaHSO₄ and β-NaHSO₄. The β-NaHSO₄ structures are mobile and move along the NaCl (110) plane until they encounter existing stationary α-NaHSO₄ structures whereupon the two forms coalescence to form larger α-NaHSO₄ structures. Gravimetry was used to determine the amount of HCl evolved upon exposure to aqueous H₂SO₄ solutions modeling atmospheric aerosols. At low H₂SO₄ concentrations, a large percentage of the HCl formed remains dissolved in the H₂O. These results indicate that for conditions simulating relative humidities above 40%, the fraction of HCl released from this reaction may be as low as 0.40 ± 0.11.

Full text of DOI:10.1021/ja001469x

J. Am. Chem. Soc. 2000, 122, 12289-12296
12289
Raman Spectroscopy and Atomic Force Microscopy of the Reaction
of Sulfuric Acid with Sodium Chloride
Christopher D. Zangmeister and Jeanne E. Pemberton*
Contribution from the Department of Chemistry, UniVersity of Arizona, 1306 East UniVersity BlVd.,
Tucson, Arizona 85721
ReceiVed April 27, 2000. ReVised Manuscript ReceiVed October 6, 2000
Abstract: The reaction of NaCl with H2SO4 is investigated using Raman spectroscopy, atomic force microscopy
(AFM), and gravimetric analysis. Raman spectra are consistent with the formation of NaHSO4 with no evidence
for Na2SO4. The spectra indicate that the phase of NaHSO4 varies with the amount of H2O in the H2SO4. At
low H2O concentrations, the reaction produces anhydrous â-NaHSO4, which undergoes a phase change to
anhydrous R-NaHSO4 over the course of 18 h. At higher H2O concentrations, anhydrous R-NaHSO4 is formed
with small amounts of NaHSO4‚H2O. AFM measurements on NaCl (100) show the formation of two distinct
types of NaHSO4 structures consistent in shape with R-NaHSO4 and â-NaHSO4. The â-NaHSO4 structures
are mobile and move along the NaCl (110) plane until they encounter existing stationary R-NaHSO4 structures
whereupon the two forms coalescence to form larger R-NaHSO4 structures. Gravimetry was used to determine
the amount of HCl evolved upon exposure to aqueous H2SO4 solutions modeling atmospheric aerosols. At
low H2SO4 concentrations, a large percentage of the HCl formed remains dissolved in the H2O. These results
indicate that for conditions simulating relative humidities above 40%, the fraction of HCl released from this
reaction may be as low as 0.40 ( 0.11.
Introduction
deficit has been attributed to reactions with inorganic oxide
species that generate volatile chlorine compounds.5-9
The majority of chlorine in the atmosphere exists in the form
of reservoir HCl.1 However, HCl can be transformed to reactive
Cl at all levels of the atmosphere. Through reaction with OH
in the troposphere, HCl is converted to Cl that can readily
abstract hydrogen from volatile organic species to form organic
radical species. In the stratosphere, the production and trans-
formation of chlorine-containing species are central to the
formation of the ozone hole, since a single Cl can result in the
destruction of ∼105 ozone molecules before one radical ClO is
terminated.1 Thus, the critical role played by reactive chlorine
species in many atmospheric processes delineates the need for
a fundamental understanding of the formation and reactivity of
HCl at all levels of the terrestrial atmosphere.
Both natural and anthropogenic sources of chlorine exist. One
natural source of tropospheric chlorine is gaseous HCl from
the reaction of atmospheric alkali halides such as NaCl with
NOx and SOx species.2,3 NaCl is added to the troposphere at a
rate of ∼1012 kg per year by the evaporation of sea salt aerosols.4
Many studies have shown a significant chloride deficit of sea
salt particles (30-80%) when compared to seawater.5-8 This
In addition, NaCl is directly injected into the stratosphere by
volcanic eruptions. For example, the El Chichon volcanic
eruption generated a plume of ∼7 wt % NaCl, and also
contained NOx10 species and aerosol droplets that were ∼80 wt
% H2SO4.11 Thus the study of the reaction of concentrated H2-
SO4 solutions with particulate NaCl is relevant to upper level
atmospheric chemistry.
Reactions of NaCl with gas-phase species which result in
release of HCl are inherently heterogeneous processes. Although
many studies of such reactions have been reported in which
the loss of gaseous reactants or the production of gaseous
products are monitored, far fewer studies have focused on the
surface chemistry involved in these reactions.2,12-14 Of those
that have been reported, many rely on the techniques of surface
science (e.g. scanning electron microscopy with energy-
dispersive X-ray spectroscopy, X-ray photoelectron spectros-
copy) which oftentimes require the use of vacuum environ-
ments12,15 that can be marginal models of atmospheric conditions.
Elegant studies of reactions on NaCl using FTIR spectroscopy
have been performed,3,13,16 but this approach can suffer from
inaccessibility to the low-frequency region of the spectrum
* To whom correspondence should be addressed.
(1) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics:
From Air Pollution to Climate Change; John Wiley and Sons: New York,
1998; p 180.
(2) Vogt, R.; Finlayson-Pitts, B. J. J. Phys. Chem. 1994, 98, 3747.
(3) Peters, S. J.; Ewing, G. E. J. Phys. Chem. 1996, 100, 14093.
(4) Graedel, T. E.; Keene, W. E. Global Biogeochem. Cycles 1995, 9,
47.
(5) Zhuang, H.; Chan, C. K.; Fang, M.; Wexler, A. S. Atmos. EnViron.
1999, 33, 4223.
(6) Kerminen, V.; Pakkanen, T. A.; Hillamo, R. E. Atmos. EnViron. 1997,
31, 2753.
(7) Harkel, M. J. Atmos. EnViron. 1997, 31, 417.
(8) Kleeman, M. J.; Hughes, L. S.; Allen, J. O.; Cass, G. R. EnViron.
Sci. Technol. 1999, 33, 4331.
(9) Poi, C. A.; Lopes, D. A. J. Geophys. Res. 1998, 103, 25263.
(10) Woods, D. C.; Chuan, R. L.; Rose, W. I. Science 1985, 230, 170.
(11) Hofmann, D. J.; Rosen, J. M. Geophys. Res. Lett. 1983, 10, 313.
(12) Allen, H. C.; Laux, J. M.; Finlayson-Pitts, B. J.; Hemminger, J. C.
J. Phys. Chem. 1996, 100, 6371.
(13) De Haan, D. O.; Finlayson-Pitts, B. J. J. Phys. Chem. A 1997, 101,
9993.
(14) Jefferson, A.; Tanner, D. J.; Eisele, F. L.; Berresheim, H. J. Geophys.
Res. 1998, 103, 1639.
(15) Laux, J. M.; Fister, T. F.; Finlayson-Pitts, B. J.; Hemminger, J. C.
J. Phys. Chem. 1996, 100, 19891.
(16) Langer, S. R.; Pemberton, R. S.; Finlayson-Pitts, B. J. J. Phys. Chem.
A 1997, 101, 1277.
10.1021/ja001469x CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/23/2000

12290 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Zangmeister and Pemberton
where many vibrational modes of simple inorganic molecules
occur and from serious spectral interference from strong
absorbers such as water.
released, gravimetry was used to measure the amount of HCl
liberated in H2SO4 solutions that model atmospheric aerosols
at various RH.
Raman spectroscopy has been used in this laboratory for
surface studies on NaCl based on its inherent molecular
specificity, its immunity from spectral interference from water,
and its routine accessibility to the low-frequency region of the
spectrum. An added benefit to this approach is that it can easily
be adapted to studies in a variety of atmospheric conditions.
When combined with atomic force microscopy (AFM) to
monitor surface morphological changes attendant to these
reactions, considerable insight into this heterogeneous chemistry
at an unprecedented level of molecular detail emerges. The
power of these complementary techniques is demonstrated here
for the study of the reaction of NaCl with H2SO4.
Experimental Section
Materials. NaCl (Aldrich, >99.99%) was recrystallized from 50:
50 H2O:EtOH and heated at 550 °C for 72 h to remove surface H2O.
This scheme results in cubes with ∼7 µm edges and a geometric surface
area of ∼4 × 103 cm2/g. H2SO4 (Aldrich, >99.999%, 4 wt % H2O)
was added to powdered NaCl at a H2SO4:NaCl of 0.5:1 at ambient
temperature and humidity (RH ranged from 7 to 25%). This experi-
mental protocol ensures that H2SO4 is the limiting reagent. Aqueous
samples of H2SO4 were prepared in sealed scintillation vials prior to
analysis and added to powdered NaCl to ensure a H2SO4:NaCl of 0.5:
1.
R-NaHSO4 was prepared by heating NaHSO4‚H2O to 120 °C for 48
h. These samples did not absorb significant amounts of H2O and did
not undergo subsequent phase changes after exposure to ambient
humidities for several weeks. â-NaHSO4 was prepared in an evacuated
flask by the addition of equal moles of Na2SO4 to 96 wt % H2SO4 at
130 °C with continual stirring for 12-24 h. The product was sealed in
glass after synthesis and was verified by XRD and Raman spectroscopy.
Instrumentation. Raman spectra were acquired using 150 mW of
532 nm radiation from a Coherent Verdi 2 Nd:vanadate laser. A Minolta
f/1.2 camera lens was used to collect scattered radiation at 90° with
respect to the incident beam. A Spex 1877 Triplemate spectrometer
coupled to a Princeton Instruments RTE-1100-PB thinned, back-
illuminated CCD camera of 1100 × 330 pixel format cooled to -90
°C was used for Raman spectroscopy.
Contact mode AFM images were acquired using a Digital Instru-
ments Multimode III SPM. A 1 × 1 cm2 piece of NaCl (100)
(International Crystal Laboratories) was cleaved in ambient (RH ∼12%)
prior to analysis. At these humidities, the NaCl (100) surface is covered
with ca. 0.20 monolayer of H2O.22 An H2SO4 aerosol was generated
by heating a solution of 96 wt % H2SO4 at ca. 120 °C in a sealed 1 cm
diameter vial. The NaCl (100) was exposed to the aerosol by placing
it over the heated vial for a prescribed period of time at 12% RH. During
exposure, the NaCl crystal effectively covered the top of the vial,
thereby preventing exposure of the solution to air. After exposure, the
surface was immediately mounted in the AFM and images acquired in
ambient.
Gravimetric experiments were performed using a Mettler Toledo
AG204 Delta Range balance with 0.1 mg sensitivity. To ensure that
NaCl was not the limiting reagent in these reactions, 2:1 NaCl:H2SO4
was used. Aqueous solutions of 0.1, 2, 4, and 8 H2O:H2SO4 were
prepared. Added to each vial was 1.07 ( 0.01 g of NaCl. Gravimetric
measurements were made in capped 20 mL scintillation vials into which
a 2 mm hole was drilled to allow HCl to escape. No significant mass
loss (<1%) was recorded 24 h after exposure. This time is comparable
to atmospheric lifetimes of NaCl aerosols in the troposphere,1 and was
used to determine the amount of HCl evolved from the reaction. The
mass of each vial was determined initially, and then at a second time
24 h after addition of the H2SO4. Control samples of H2O and NaCl
were prepared to account for mass loss from H2O evaporation. Ten
replicate samples were acquired for each set of exposures.
The nucleation, sources, and sinks of H2SO4 in the atmosphere
have been extensively studied.14,17 H2SO4 is extremely hygro-
scopic and its vapor pressure is 106 times lower than that of
H2O. Therefore, unlike other mineral acids, H2SO4 is found as
a hydrated liquid in the atmosphere. The reaction of H2SO4 with
5,18-21
NaCl has been proposed
to result in the formation of
Na2SO4 with release of two moles of HCl per mole of H2SO4:
2NaCl (s) + H2SO4 (l) a 2HCl (g) + Na2SO4 (s) (1)
Analysis of atmospheric particles from the El Chichon eruption
by energy-dispersive X-ray (EDX) spectroscopy showed the
presence of a sulfur signal attributed to Na2SO4.10 Although
EDX spectroscopy is sensitive to the element of interest, it is
insensitive to either the oxidation or protonation states of species
from which the signals arise. Therefore, the chemical nature of
the species from which these S signals arose cannot be
determined. In more recent work, ten Brink studied the reaction
of NaCl with H2SO4 in a smog chamber and found that the
reaction kinetics were dependent on the size of the NaCl particle,
but he was also unable to prove a Na2SO4 product.19
Understanding reaction pathways of atmospheric reactions
is fundamental in the assessment of the chlorine budget.
Although collection and analysis of atmospheric aerosols and
particles provides information on the final products of such
reactions, this information only leads to speculation about the
relevant reaction mechanisms.
We have studied the reaction of NaCl with H2SO4 using
Raman spectroscopy, a technique that allows direct interrogation
of the products formed and their crystallographic phases.
Mechanistic insight into the pathways by which these phases
form on the NaCl surface can be ascertained from atomic force
microscopy.
H2O is ubiquitous in the terrestrial atmosphere and greatly
affects the reaction probabilities of NaCl with NOx species, but
its effect on SOx chemistry is unknown. Given the hygroscopic
nature of H2SO4, atmospheric H2SO4 will be hydrated even at
low relative humidities (RH). We have modeled this effect by
exposing particulate NaCl to solutions of varying H2O:H2SO4
ratios.
Results and Discussion
Raman Spectroscopy. Differences in molecular symmetry
between HSO4- and SO42- result in distinct Raman spectra for
different crystalline solids of the sodium salts.23-25 As shown
by the spectra in Figure 1 and the tabulated peak frequencies
in Table 1, Na2SO4 has major bands at 467, 633, 994, and 1132
cm-1 while NaHSO4‚H2O has bands at 439, 606, 878, and 1042
Models of this chemistry generally assume that all of the
chlorine produced as HCl in this reaction is liberated to the
atmosphere.5,18-21 To better quantify the amount of HCl
(17) Weber, R. J.; McMurry, P. H.; Mauldin, R. L.; Tanner, D. J.; Eisele,
F. L.; Clarke, A. D.; Kapustin, V. N. Geophys. Res. Lett. 1999, 26, 307.
(18) Hitchcock, D. R.; Spiller, L. L.; Wilson, W. E. Atmos. EnViron.
1980, 14, 165.
cm-1 26-28
R-NaHSO4 has clusters of major bands around 430
.
and 600 cm-1 with additional prominent bands at 869, 1006,
(19) ten Brink, H. M. J. Aerosol Sci. 1998, 29, 57.
(22) Peters, S. J.; Ewing, G. E. Langmuir 1997, 13, 6345.
(20) Jefferson, A.; Eisele, F. L.; Ziemann, P. J.; Weber, R. J.; Marti, J.
J.; McMurry, P. H. J. Geophys. Res. 1997, 102, 19021.
(21) Symonds, R. B.; Rose, W. I.; Reed, M. H. Nature 1988, 334, 415.
(23) Choi, B.; Lockwood, D. J. Solid State Commun. 1989, 72, 133.
(24) Sonneveld, E. J.; Visser, J. W. Acta Crystallogr. 1979, B35, 1975.
(25) Sonneveld, E. J.; Visser, J. W. Acta Crystallogr. 1978, B34, 643.

Reaction of Sulfuric Acid with Sodium Chloride
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12291
Figure 1. Raman spectra of bulk (a) Na2SO4, (b) NaHSO4‚H2O, (c)
anhydrous R-NaHSO4, and (d) anhydrous â-NaHSO4.
Figure 2. Solution Raman spectra of (a) 96 wt % H2SO4, (b) saturated
aqueous NaHSO4‚H2O, and (c) saturated aqueous Na2SO4.
Table 1. Raman Frequencies of Sulfate Salts
peak freq (cm-1
R-NaHSO4 â-NaHSO4 NaHSO4‚H2O Na2SO4 assignmenta,b
)
416
427
447
469
568
587
605
630
869
421
417
439
578
606
δ(S-O)
δ(S-O)
δs(SO3)
439
451
613
981
575
591
δa(SO3)
617
904
858
878
1042
νs(S-O-H)
νs(S-O)
νa(S-O)
1006
1066
1250
1031
NOc
1243
1097
1101
1152
a Assignments for R-NaHSO4, â-NaHSO4, and NaHSO4‚H2O from
refs 26, 27, and 28; assignment for Na2SO4 from ref 23. b δ ) bend;
ν ) stretch; s ) symmetric; a ) asymmetric. c Not observed in these
experiments.
Figure 3. Raman spectra of 96 wt % H2SO4:NaCl at 0.5:1 after a
reaction time of (a) 5 min, (b) 6 h, and (c) 24 h.
1066, and 1250 cm-1 28
430 and 580 cm-1 and additional major bands at 904 and 1031
cm-1 28
The most intense peak in the spectra of all forms is the
νs(SO4). For Na2SO4, this mode is a sharp singlet at 994 cm-1
.
â-NaHSO4 has clusters of bands at ∼
which appears at 914 cm-1 in 96 wt % H2SO4, shown in Figure
2a. The spectrum of a saturated aqueous solution of NaHSO4‚
H2O shown in Figure 2b has two intense modes at 981 and
.
2-
1055 cm-1 which are assigned to the νs(SO4) of SO4 and
HSO4-, respectively.29 The solution of saturated Na2SO4 is
.
dominated by the νs(SO4) at 981 cm-1
.
In the spectrum of R-NaHSO4, this mode is a doublet with peaks
at 1006 and 1066 cm-1 due to the presence of two discrete
Raman spectra were acquired with time after exposure of
powdered NaCl to 96 wt % H2SO4 at a H2SO4:NaCl ratio of
0.5:1 at 25% RH. The spectra shown in Figure 3a-c represent
spectra of the reaction products 5 min, 6 h, and 24 h after
reaction, respectively. These spectra reflect a slow, systematic
change in NaCl surface product over this period. None of these
spectra is consistent with the presence of Na2SO4, the previously
assumed5,18-21 product of this reaction. Instead, the observation
of a νs(SO4) mode at frequencies greater than 1000 cm-1
suggests a NaHSO4 product. The observation of a distinct vs-
(S-O-H), initially at 904 cm-1 but shifting to 869 cm-1 with
time, is further consistent with NaHSO4.
-
HSO4 environments in the unit cell.25,26 â-NaHSO4 has a
crystal structure in which all HSO4- are equivalent and exist in
dimerized pairs.25 Thus, the νs(SO4) of â-NaHSO4 occurs as a
single peak at 1031 cm-1. NaHSO4 spectra have an additional
band in the vicinity of 870-910 cm-1 due to the νs(S-O-H).
2-
-
Raman spectra of aqueous SO4 and HSO4 solutions are
shown in Figure 2. These spectra have bands which are
significantly broader than those in the spectra of the crystalline
solids. The most intense mode in each spectrum is the νs(SO4),29
(26) Dey, B.; Jain, Y. S.; Verma, A. L. J. Raman Spectrosc. 1982, 13,
209.
(27) Goypiron, A.; de Villepin, J.; Novak, A. J. Raman Spectrosc. 1980,
9, 297.
(28) Baran, J.; Ilczszyn, M. M.; Marchewka, M. K.; Ratajczak, H.
Spectrosc. Lett. 1999, 32, 83.
The spectrum in Figure 3a for the reaction product formed
immediately after exposure of NaCl to H2SO4 does not match
that of either R-NaHSO4 or NaHSO4‚H2O as shown in Figure
1, spectra b and c, respectively. In contrast, the spectrum in
(29) Irish, D. E.; Chen, H. J. Phys. Chem. 1970, 74, 3796.

12292 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Zangmeister and Pemberton
Figure 4. H2O:H2SO4 as a function of RH (neglecting the Kelvin
effect). Arrows indicate RH values modeled in this study.
Figure 3a can be attributed to â-NaHSO4, a relatively unstable
form of NaHSO4.25,28
Figure 5. Raman spectra of NaCl after reaction with (a) 2 H2O:H2-
SO4, (b) 4 H2O:H2SO4, (c) 8 H2O:H2SO4, (d) 8 H2O:H2SO4 after 24 h,
and (e) 20:1 H2O:H2SO4.
Previous researchers have reported considerable difficulty in
isolating pure â-NaHSO4 in the presence of gas-phase mois-
ture.25 Indeed, we have recently confirmed that H2O catalyzes
the transformation of pure bulk â-NaHSO4 to R-NaHSO4.30 The
same transformation occurs for NaCl surface-confined â-NaH-
SO4 formed in the reaction with H2SO4. Within 10 min of H2-
SO4 exposure, the νs(S-O-H) at 904 cm-1 decreases in
intensity at the expense of a new νs(S-O-H) at 869 cm-1
(Figure 3b,c). Growth of two new νs(SO4) modes at 1011 and
1062 cm-1 at the expense of that at 1031 cm-1 is also consistent
with the transformation from â-NaHSO4 to R-NaHSO4.
This transformation takes ∼15-18 h at the relative humidities
(7-25%) used in this study. Given that the atmospheric lifetime
of NaCl particles is on the order of 6-24 h depending on size,
these changes may manifest themselves on atmospheric particles
prior to deposition.
hydrates,31 and they model a range of humidities in the terrestrial
atmosphere. In fact, the hydrate formed in the 4:1 mixture has
been observed in the atmosphere.33 RH data from coastal sites
within the United States were obtained from the National
Weather Service (NWS) and National Oceanic and Atmospheric
Administration (NOAA).34 The 24 h average RH obtained from
77 coastal sites located in the United States and its territories is
73 ( 11%. Assuming that the evaporation of sea salt aerosols
occurs in proximity to the coast, one would predict that H2SO4
exists as an aqueous aerosol with a composition of 20:1. Thus,
these compositions, particularly those of the hydrates with larger
amounts of H2O, are atmospherically relevant in the marine
boundary layer.
Spectra of the products formed by the exposure of powdered
NaCl to H2SO4 solutions such that the final mixtures possessed
compositions of NaCl:H2O:H2SO4 of 2:2:1, 2:4:1, 2:8:1, and
2:20:1 are shown in Figure 5. Each spectrum was acquired ∼15
min after exposure to the H2SO4 solution. The reaction of NaCl
with a 2:1 H2O:H2SO4 solution results in the formation of
R-NaHSO4 as shown by the spectrum in Figure 5a. However,
after 24 h, this â-NaHSO4 has undergone transformation to
R-NaHSO4.
Several experiments were also undertaken in which powdered
NaCl was exposed to H2SO4 by passing a stream of aerosol
droplets created by evacuation of 96 wt % H2SO4 over the NaCl.
This experimental arrangement better reflects the exact nature
of atmospheric exposure of NaCl to H2SO4 which exists as
aerosol droplets. The spectral results indicate the formation of
R-NaHSO4 after ∼ 48 h of H2SO4 aerosol exposure (data not
shown). â-NaHSO4 was not observed, probably due to small
amounts of H2O vapor in the headspace which cause the phase
transition on this time scale.
The Raman spectrum in Figure 5b shows the product of
exposing NaCl to 4:1 H2O:H2SO4 to be R-NaHSO4 based on
H2SO4 that is 96 wt % has a H2O:H2SO4 ratio of 1:9. This
composition is a poor model for H2SO4 in the terrestrial
atmosphere where the water content is much higher. The
hydration of H2SO4 in the atmosphere depends on RH; several
stable hydrates of H2SO4 are known to form.31 Figure 4 shows
a plot of the H2O:H2SO4 ratio in these hydrates at 298 K as a
function of RH1,32 (neglecting the Kelvin effect.) The hydration
of H2SO4 increases linearly between a RH of 0 and 70%. Above
this point, the hydration number increases exponentially to
nearly 30 at ∼80% RH.
the frequencies of the νs(SO4) doublet at 1006 and 1066 cm-1
.
In addition, the weak νs(SO4) at 1042 cm-1 is consistent with
the formation of a small amount of NaHSO4‚H2O. This product
is a solid that is visibly wet. Spectroscopic evidence for the
presence of surface H2O is the shoulder at 981 cm-1 which is
attributed to aqueous HSO4- species as shown by the spectrum
of saturated NaHSO4‚H2O solution in Figure 2b. Spectra taken
24 h after exposure to this H2SO4 hydrate (not shown) indicate
the loss of the νs(SO4) mode of NaHSO4‚H2O at 1042 cm-1
indicating the conversion of all NaHSO4‚H2O to anhydrous
R-NaHSO4. Given that this sample is a highly viscous slurry,
the transformation to a completely anhydrous bisulfate product
is surprising.
To model environments of 10, 22, 55, and 85% RH, solutions
of 2:1, 4:1, 8:1, and 20:1 H2O:H2SO4 mixtures, respectively,
were prepared. These compositions were chosen because they
have been shown to be thermodynamically stable H2SO4
Exposure of NaCl to 8:1 H2O:H2SO4 yields the spectrum
(30) Zangmeister, C. D.; Pemberton, J. E. J. Solid State Chem. Submitted
for publication.
shown in Figure 5c. The bands at 1006, 1042, and 1066 cm-1
(31) Giaque, W. F.; Hornung, E. W.; Kunzler, J. E.; Rubin, T. R. J. Am.
Chem. Soc. 1959, 82, 62.
(32) Liu, B. Y. H.; Levi, J. In Generation of Aerosols and Facilities for
Exposure Experiments; Willeke, K., Ed.; Ann Arbor Science: Ann Arbor,
MI, 1980; pp 317-336.
(33) Sassen, K.; Peter, T.; Luo, B. P.; Crutzen, P. J. Appl. Opt. 1994,
33, 4602.
(34) Obtained from relative humidity data collected by the National
Oceanographic and Atmospheric agency (NOAA) and the National Weather
Service (NWS) from http://met-www.cit.cornell.edu/ccd/relhum98.html.

Reaction of Sulfuric Acid with Sodium Chloride
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12293
are consistent with the presence of both R-NaHSO4 and
NaHSO4‚H2O. The partial dissolution of solid NaHSO4 is more
evident in the 8:1 sample as shown by the more pronounced
presence of the solution HSO4- 981 cm-1 mode. Spectra taken
24 h after exposure (Figure 5d) indicate a dramatic reduction
in the amount of NaHSO4‚H2O at the expense of an increase in
R-NaHSO4 modes.
To simulate 85% RH, a 20:1 H2O:H2SO4 solution was
prepared. This sample contains ∼10 mol of H2O per mol of
NaCl. Thus, this sample contains sufficient H2O to entirely
dissolve the solid. The Raman spectrum shown in Figure 5e is
of the resulting solution product. This spectrum is similar to
that of saturated aqueous NaHSO4 shown in Figure 2b.
2-
However, the intensity of the 984 cm-1 mode of solution SO4
is less intense relative to the 1053 cm-1 mode of solution HSO4
-
reflecting a lower solution pH than in saturated aqueous
NaHSO4. This lower pH is the result of HCl production in this
reaction. Assuming total conversion of NaCl to NaHSO4, the
HCl concentration in this solution is ∼1.4 M, although Brimble-
combe et al.35 have shown that HCl degassing is thermodynami-
cally favored at pH <3. Therefore, the initial HCl generated
from this reaction is degassed until the equilibrium value of
Figure 6. Raman spectra of â-NaHSO4 on NaCl exposed to 100%
RH. Spectra of (a) â-NaHSO4 prior to exposure, (b) mixture of
â-NaHSO4 and R-NaHSO4 10 min after exposure, (c) pure R-NaHSO4
20 min after exposure, (d) mixture of NaHSO4‚H2O and R-NaHSO4
30 min after exposure, and (e) pure NaHSO4‚H2O 40 min after exposure.
2-
pH ∼3 is reached, a pH at which partial SO4 protonation to
-
HSO4 is favored.
The presence of H2O in the H2SO4 does not change the
product of the NaCl reaction from NaHSO4; however, the phase
of anhydrous NaHSO4 formed depends on the amount of H2O
available for hydration of the sample. It is possible to form
NaHSO4‚H2O at large relative amounts of H2O. However, the
NaHSO4‚H2O is not stable on the NaCl surface and loses H2O
to become the anhydrous R-NaHSO4 phase over the course of
24 h. The loss of NaHSO4‚H2O with the concomitant growth
of R-NaHSO4 may be due to hydration of NaCl at the expense
of HSO4-. The crystallographic rearrangement required for such
a transformation must be kinetically slow, and thus require up
to ∼24 h to achieve equilibrium. It is unlikely that H2O
evaporation results in the observed changes, as the samples were
kept tightly sealed between and during sample acquisitions.
The apparent desiccating effect of NaCl in this system
suggests greater affinity for H2O by NaCl than by R-NaHSO4.
Given that the deliquescence points (DP) of NaCl and NaHSO4‚
H2O are 74.5% and 52%, respectively,1 this observation may
seem somewhat surprising. However, in this system, it is the
relative affinity of H2O by NaCl and R-NaHSO4 that is relevant,
and since the DP of R-NaHSO4 cannot be determined due to
its conversion to NaHSO4‚H2O, this relative affinity cannot be
known. Based on the observation that NaCl takes H2O from
NaHSO4‚H2O on its surface, one must conclude that either the
presence of the NaCl surface raises the DP of NaHSO4‚H2O
above that of NaCl or the DP of R-NaHSO4 is higher than that
of either NaCl or NaHSO4‚H2O. The atmospheric relevance of
these observations is that the presence of unexpected products
in such reactions may significantly alter the expected uptake of
H2O by solid particles.
spectral contributions from processes such as surface deliques-
cence are not observed in this study.
To further study the effect of H2O on the phase transition of
anhydrous â-NaHSO4 to R-NaHSO4, several samples of NaCl
were exposed to 96 wt % H2SO4 (0.5:1 H2SO4:NaCl) in a dry
bag and stored in a desiccator for several days. Spectra of these
samples indicate the initial formation of pure â-NaHSO4 in this
dry environment. Even after 4 days, the â-NaHSO4 formed in
this reaction is stable if it is kept dry as indicated by the spectrum
in Figure 6a.
Exposure of this 1:1 â-NaHSO4:NaCl sample to an environ-
ment of 100% RH produces the sequence of spectra shown in
Figure 6b-e. Figure 6b was acquired ∼10 min after exposure
to the 100% RH environment and indicates growth of the
R-NaHSO4 phase. Further exposure to 100% RH results in the
spectrum shown in Figure 6c. At this point, nearly all of the
â-NaHSO4 has been converted to R-NaHSO4. As this powder
is continually exposed to 100% RH, the νs(SO4) mode of
NaHSO4‚H2O at 1042 cm-1 appears as shown in Figure 6d.
Figure 6e was acquired ∼30 min after Figure 6b; at this point,
the conversion of R-NaHSO4 to NaHSO4‚H2O is complete. Very
little spectral evidence for the presence of solution phase species
exists in Figure 6b-e. This absence may be due to the fast
uptake of H2O by R-NaHSO4 to form NaHSO4‚H2O and/or a
R-NaHSO4 DP greater than the 52% DP of NaHSO4‚H2O.
Collectively, these results clearly indicate that the presence
of H2O catalyzes the transformation of anhydrous â-NaHSO4
to anhydrous R-NaHSO4. H2O-mediated phase transitions have
been observed for other HSO4- salts;36-39 however, this is the
first such observation for NaHSO4. Careful inspection of the
crystal structures of these three phases indicates that those of
R-NaHSO4 and NaHSO4‚H2O are similar25,28 with each having
two discrete HSO4- environments per unit cell.24,28 We speculate
Unlike the reaction of NaCl with HNO3 and NO2,13 a thin
passivating surface layer of the reaction product is not observed
in this reaction. The species present in this study (NaCl and
NaHSO4) are optically transparent at the excitation wavelength
used; therefore, the sampling depth is large and extends much
beyond the surface into the bulk. Indeed, the sampling depth
may exceed the mean powder size (∼7 µm) depending on
scattering. The signal arising from the surface of the sample is
much less than the Raman scattering from the bulk. Thus,
(36) Baranowski, B.; Lipkowski, J.; Lunde´n, A. J. Solid State Chem.
1995, 117, 412.
(37) Baranowski, B.; Friesel, M.; Lunde´n, A. Z. Naturforsch. 1987, 42A,
565.
(38) Baranowski, B.; Friesel, M.; Lunde´n, A. Z. Naturforsch. 1986, 41A,
981.
(39) Colomban, Ph.; Pham-Thi, M.; Novak, A. Solid State Ionics 1987,
24, 193.
(35) Brimblecombe, P.; Clegg, S. L. J. Atmos. Chem. 1988, 7, 1.

12294 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Zangmeister and Pemberton
Figure 7. Contact AFM images after exposure of NaCl (100) to H2SO4 aerosol: (a, b) images 4 min after exposure; (c, d) images 12 min after
exposure; (e, f) images 20 min after exposure; (a, c, and e) 5 × 5 µm2 region; (b, d, and f) 1 × 1 µm2 region.
Table 2. Unit Cell Dimensions of Anhydrous NaHSO4 Phases
that the phase transition from â-NaHSO4 to R-NaHSO4 is
initiated by hydration of Na+ in the â phase. Exposure to higher
RH results in the incorporation of H2O into the crystal structure
to form NaHSO4‚H2O. We have recently proven that the
â-NaHSO4 to R-NaHSO4 phase transition also occurs in the
absence of NaCl, but much more slowly.30 Thus, one can
speculate that NaCl facilitates H2O uptake, thereby increasing
the rate of the phase transition.
unit cell parametersa
phase
a (Å)
b (Å)
c (Å)
R-NaHSO4
â-NaHSO4
7.005
8.759
7.125
7.500
6.172
5.147
a References 24 and 25.
Atomic Force Microscopy. Further insight into the initial
stages of this reaction comes from AFM on NaCl (100) surfaces
exposed to H2SO4 aerosols. The average diameters of airborne
H2SO4 aerosol droplets and NaCl particles have been estimated
to be ∼1 and 5 µm, respectively.1 Thus, to create an atmo-
spherically relevant model of the interaction of H2SO4 aerosols
with the surfaces of NaCl particles, H2SO4 droplets that are
smaller than the NaCl surface are desired. Such an aerosol
population was achieved by heating 96 wt % H2SO4 in a closed
vial at 120 °C at 12% RH. Based on the data in Figure 4, these
H2SO4 aerosols are assumed to be composed of ∼3:1 H2O:H2-
SO4. Subsequent handling and mounting of the NaCl crystal
after exposure takes additional time. Thus, scanning was initiated
∼4 min after exposure of the NaCl surface to these H2SO4
aerosols. It should also be noted that the surface was exposed
to 12% RH during the entire exposure and AFM analysis
sequence.
A series of real-time contact mode AFM images collected in
one region of the surface over 16 min is shown in Figure 7.
Vertical stripes that run down the length of the images in Figure
7a,c,e are monatomic steps on the NaCl surface. A uniform
distribution of reaction products is observed that is consistent
with reaction of aerosol droplets with the NaCl surface.
After 5 min of H2SO4 aerosol exposure (Figure 7b), station-
ary, pyramidal-shaped structures that are separated by ∼100
nm form at step edges and on terrace sites. Assuming that these
structures are NaHSO4, their size and aspect ratio provide insight
into which phase they represent. The dimensions of these
structures are 55.7 ( 8.0 nm × 58.9 ( 12.7 nm (x-y aspect
ratio of 1.1 ( 0.1) with an average height of ∼20 nm. This
ratio compares favorably to the 1.15 b/c aspect ratio of
R-NaHSO4 as shown by the crystallographic data in Table 2.25
Thus, based on this aspect ratio and the Raman spectral data of
this reaction shown in Figure 3c, these structures are proposed
to be R-NaHSO4.
In addition to these R-NaHSO4 particles, a large population
of smaller structures (6-8 times as many as the R-NaHSO4
particles) is also observed on the surface. In Figure 7a,b, these
structures have dimensions of 12.9 ( 1.9 nm × 22.4 ( 3.2 nm
(x-y aspect ratio of 1.7 ( 0.2) with an average height of 2.5
nm. This aspect ratio compares favorably to the 1.70 a/c aspect
ratio of â-NaHSO4 as shown in Table 2.24 On the basis of this
aspect ratio and the Raman spectral data shown in Figure 3a,
these smaller structures are proposed to be â-NaHSO4.
Continuous scanning of the same surface region results in
the images shown in Figure 7c-f. In Figure 7c,d, many of the
smaller â-NaHSO4 structures are observed to be mobile across
the surface. These â-NaHSO4 structures disappear from the
surface by the next scan (Figure 7e taken ∼4 min later) and
the nearby R-NaHSO4 structures are larger in Figure 7e than in

Reaction of Sulfuric Acid with Sodium Chloride
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12295
Figure 7a. The â-NaHSO4 structures are mobile and move at
45° with respect to the NaCl (100) plane, or in other words,
along the NaCl (110) plane. Motion of â-NaHSO4 along the
NaCl (110) direction must represent the lowest energy plane of
motion. This movement can be rationalized by realizing that
the a/c plane of â-NaHSO4 has Na+ in the plane24,25 coincident
with the NaCl surface that allows movement along a single row
of Na+ without encountering Cl- as would occur for movement
along the NaCl (100) plane.
The movement of â-NaHSO4 ceases over the 16 min that
span the data in Figure 7. The encounter of â-NaHSO4 with
R-NaHSO4 results in coalescence of the two structures. As a
result, the R-NaHSO4 structures increase in size while still
retaining their 1.1 ( 0.1 aspect ratio.
In some cases, movement of the â-NaHSO4 structures results
in encounters with a step edge before an existing particle of
R-NaHSO4. In ∼95% of such cases, the â-NaHSO4 structures
at the step edge coalesce and transform into a single R-NaHSO4
structure based on their aspect ratios. This transformation is
probably the result of the presence of H2O, which collects
preferentially at the step edges,40 that mediates this transforma-
tion as noted above. The remaining 5% of these encounters result
in crossing of the step edge by the â-NaHSO4 and continued
movement along the NaCl (110) plane.
â-NaHSO4 surface translation is hypothesized to be promoted
either by dissolution and reprecipitation of NaHSO4 or cation-
anion place exchange along the (110) plane. H2O has been
observed to facilitate the rearrangement of soluble surfaces in
previous work,40,41 and may be responsible for surface transla-
tion in this study. Since H2O initiates the phase transition from
â- to R-NaHSO4, any NaHSO4 that was to reprecipitate would
be R-NaHSO4. Thus, dissolution and reprecipitation is not a
feasible mechanism for translation of â-NaHSO4, leaving the
ion place exchange as the most likely translation mode. H2O
may aid this place exchange by dissolution of the NaCl surface
during the process.
The transformation of anhydrous â-NaHSO4 to R-NaHSO4
is mediated by H2O and the rate at which the transition takes
place on the NaCl surface is greatly influenced by the RH. This
observation explains the difference in time for the phase
transition as monitored by Raman spectroscopy in a closed
environment as compared to that monitored in the ambient AFM
experiments. By carefully controlling the amount of H2O in these
experiments, the phase transition can be totally eliminated for
up to several days.
Gravimetry of HCl Release. The hygroscopic nature of H2-
SO4 ensures that the acid is always hydrated in the earth’s
atmosphere. The presence of water may have two effects in the
reaction of NaCl with H2SO4. First, H2O might increase the
kinetics of the reaction as has been observed in the reaction of
NaCl with nitrogen oxide species.13 In addition, H2O may also
sequester the HCl produced in this reaction by dissolution. Thus,
the amount of HCl that is released into the atmosphere may be
considerably less than that produced in the reaction. Previous
studies on the release of atmospherically relevant gas-phase acids
in solution have been undertaken.35,42 However, these studies
were confined to reactions that occur in droplets or solution.
The work reported here is the first such study of the release of
surface-confined HCl from a H2O-limited environment.
To determine the mass of HCl released as a function of the
H2O content of the H2SO4, a series of gravimetric experiments
Figure 8. (a) Percent HCl released from NaCl surface as a function
of H2O:H2SO4; (b) HCl release per H2O as a function of H2O:H2SO4.
Table 3. Gravimetric Results for HCl Evolution as a Function of
H2O:H2SO4
% HCl lost of total
H2O:H2SO4 soln
% RH simulateda
HCl producedb
0.1:1
2:1
4:1
<1
10
22
55
77.4 ( 4.5
72.6 ( 9.4
40.2 ( 11.1
46.8 ( 9.9
8:1
a References 1, 31,and 32. b Based on mass loss over 24 h.
were performed in which the total system mass was monitored
before reaction and after reaction and equilibration of the
products with the surrounding atmosphere in an open vessel.
Once reaction and equilibration had occurred, no further mass
changes were noted for periods up to 24 h. The significance of
these experiments is that they represent the first studies done
under tropospherically relevent pressures. Previous studies of
HCl release18 have been performed in vacuum environments
which facilitate removal of the HCl formed in the reaction.
These previous experiments performed under vacuum19 might
be expected to provide results different from those that occur
at tropospheric pressures.
Four aqueous solutions of H2SO4 were prepared with H2O:
H2SO4 of 0.1, 2, 4, and 8. These solutions were added to
powdered NaCl in a 2:1 NaCl:H2SO4 solution ratio. The
reactions in 0.1 and 2 H2O:H2SO4 solutions produce solid
products, whereas those in 4 and 8 H2O:H2SO4 solutions result
in products containing both solid and liquid phases. The percent
HCl released from the solid product as a function of H2O:H2-
SO4 is shown in Figure 8a and tabulated in Table 3. These results
indicate that the solid products release ∼30% more HCl than
the mixture of solid and liquid-phase products.
The effect of the presence of H2O on the release of HCl from
the solid products is pronounced. As shown in Figure 8b, ∼0.68
HCl per H2O is released for 0.1 H2O:H2SO4. This value drops
to 0.12 HCl per H2O at 2 H2O:H2SO4 and further decreases to
0.025 for 4 and 8 H2O:H2SO4. Thus, as the amount of H2O in
(40) Dui, Q.; Hu, J.; Salmeron, M. J. Phys. Chem. B 1997, 101, 1994.
(41) Zangmeister, C. D.; Pemberton, J. E. J. Phys. Chem. B 1998, 102,
8950.
(42) Clegg, S. L.; Brimblecombe, P. Atmos. EnViron. 1988, 22, 91.

12296 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Zangmeister and Pemberton
the H2SO4 increases, the amount of HCl released decreases as
there is more H2O available for dissolution of the HCl produced.
Atmospherically, dissolution of HCl may significantly decrease
the amount of HCl released at humidities greater than ∼30%.
In light of the work by Brimblecombe and Clegg on bulk
aqueous solutions,35,42 these results are not too surprising.
However, the significance of these observations lies in the fact
that the behavior observed in ultrathin H2O films on NaCl is
essentially identical to that observed in bulk aqueous solutions.
Atmospheric Relevance. The El Chichon volcanic eruption
generated ∼3 × 107 molecules of 60-80 wt % H2SO4 per
cm-3 11 and 5.2 × 109 molecules of particulate NaCl per cm-3
in the stratosphere.43 Furthermore, recent work has shown that
H2SO4 aerosols exist in the liquid state at stratospheric
temperatures.44-46 Considering the number density of the each
of the reactants, the release of HCl from the reaction of NaCl
with H2SO4 aerosols is likely at these high altitudes.
HNO3.9 Thus, the extent of the NaCl reaction with H2SO4 may
also be high in such an environment. The formation of only 1
mol of HCl per mol of H2SO4 reduces the impact of this reaction
on the generation of chlorine reservoir species.
Conclusions
The reaction of NaCl with H2SO4 has been investigated as a
function of H2O content using Raman spectroscopy, AFM, and
gravimetric analysis. The product formed is NaHSO4, not Na2-
SO4 as previously proposed, and is independent of H2O content.
The initial product formed is anhydrous â-NaHSO4 which
undergoes an H2O-mediated phase transition to the anhydrous
R-NaHSO4 phase. The kinetics of this transition are extremely
sensitive to RH.
The â- to R-NaHSO4 phase transition is also observed using
AFM on NaCl (100). Exposure of NaCl (100) to H2SO4
produces NaHSO4 structures with aspect ratios consistent with
â-NaHSO4. These species are mobile on the NaCl (100) surface,
moving exclusively along the (110) plane. Encounter of these
â-NaHSO4 structures with an existing R-NaHSO4 structure or
a step edge leads to their transformation to R-NaHSO4.
Gravimetric analysis of the reaction of NaCl with H2SO4
indicates that the amount of HCl released is dependent on the
amount of H2O available for dissolution of the product. The
fraction of HCl released is reduced at high RH by up to ∼40%.
The loss of Cl- from sea salt particles has been measured at
many sites across the earth’s surface.5-9 For example, Poi et
al. observed at a site in Portugal Cl- depletions greater than
85%, with a depletion rate of 10% Cl- per h.9 In such an
atmosphere, SO2 levels, a H2SO4 precursor,47 are 4.3 times
greater than those of NO2 and 5.7 times higher than those of
(43) Michelangeli, D. V.; Allen, M.; Yung, Y. L. Geophys. Res. Lett.
1991, 18, 673.
(44) Bertram, A. K.; Patterson, D. D.; Sloan, J. J. J. Phys. Chem. 1996,
100, 2376.
(45) Martin, S. T.; Salcedo, D.; Molina, L. T.; Molina, M. J. J. Phys.
Chem. B 1997, 101, 5307.
(46) Koop, T.; Ng, H. P.; Molina, L. T.; Molina, M. J. J. Phys. Chem.
A 1998, 102, 8924.
(47) Jefferson, A.; Tanner, D. J.; Eisele, F. L.; Berresheim, H. J. Geophys.
Res. 1998, 103, 1639.
Acknowledgment. The authors gratefully acknowledge the
National Science Foundation (CHE-9504345 and CHE-0075813)
for financial support of this research.
JA001469X