An analytical approach for a comprehensive study of organic aerosols

Article abstract of DOI:10.1002/1521-3773(20011105)40:21<3998::AID-ANIE3998>3.0.CO;2-D

Unknown products identified precisely: The coupling of liquid chromatography to NMR spectroscopy, mass spectrometry, and infrared spectroscopy and the use of high-resolution mass spectrometry is utilized to investigate the formation of atmospheric-relevan

Full text of DOI:10.1002/1521-3773(20011105)40:21<3998::AID-ANIE3998>3.0.CO;2-D

COMMUNICATIONS
After washing with dry dichloromethane (10 mL), the compound
1
An Analytical Approach for a Comprehensive
Study of Organic Aerosols**
(0.125 mmol) and a catalytic amount of TEMPO (0.1 mol%) in dry
dichloromethane (10 mL) were pumped in a recycle mode under argon for
6 h through the reactor. The reactor system was rinsed with dichloro-
methane (10 mL) and the combined organic solvents were removed in
vacuo and the residue was used for the further reactions without additional
purification.
Wolfgang Schrader,* Jutta Geiger,
Markus Godejohann, Bettina Warscheid, and
Thorsten Hoffmann
Received: May 17, 2001 [Z17127]
Since the first scientific explanation of the so-called ªblue
hazeº phenomenon above forests by Went,[1] the formation of
natural organic aerosol particles in the atmosphere and their
impact on the climate has drawn considerable attention.
Airborne particles absorb, reflect, and scatter incoming solar
radiation, play an important role in cloud droplet formation,
and may even be involved in multiphase atmospheric
chemistry.[2]
In recent reports it was established that especially the
reaction of mono- and sesquiterpenes with tropospheric
ozone contributes to the formation of organic aerosols,[3]
however, our knowledge about reaction mechanisms leading
to low-volatile products is still limited. Making things difficult
is the fact that during the ozonolysis free OH radicals are
formed that have to be considered.[4]
However, the initial steps of the terpene ozonolysis in the
gas phase corresponds to known mechanisms,[5] exemplified in
Scheme 1 for a-pinene 1, the most abundant single mono-
terpene released by plants. The first step is the formation of
the molozonide 2 that forms reactive Criegee intermediates
3a and 3b. Subsequent reactions lead to the formation of
several low vapor-pressure products, such as pinonic acid (4),
pinic acid (5), or hydroxy pinonic (6) acid which have been
identified using mass spectrometric methods.[6] Although
possible mechanisms for the generation of these carboxylic
products have been suggested,[7] the overall understanding of
the reaction mechanisms leading to condensable species is still
inadequate.
[1] Recent reviews: a) A. Kirschning, H. Monenschein, R. Wittenberg,
Angew. Chem. 2001, 113, 670 ± 701; Angew. Chem. Int. Ed. 2001, 40,
650 ± 679; b) A. Kirschning, H. Monenschein, R. Wittenberg, Chem.
Eur. J. 2000, 6, 4445 ± 4450; c) S. V. Ley, I. R. Baxendale, R. N. Bream,
P. S. Jackson, A. G. Leach, D. A. Longbottom, M. Nesi, J. S. Scott, R. I.
Storer, S. J. Taylor, J. Chem. Soc. Perkin Trans. 1 2000, 3815 ± 4195;
d) D. H. Drewry, D. M. Coe, S. Poon, Med. Res. Rev. 1999, 19, 97 ± 148.
[2] ªPolymer Reagents and Catalystsº: R. T. Taylor, ACS Symp. Ser. 1986,
308, 132 ± 154.
[3] Reviews: a) D. Hudson, J. Comb. Chem. 1999, 1, 333 ± 360; D. Hudson,
J. Comb. Chem. 1999, 1, 403 ± 457; b) A. R. Vaino, K. D. Janda, J.
Comb. Chem. 2000, 2, 579 ± 596; c) P. Hodge, Chem. Soc. Rev. 1997, 26,
417 ± 424.
[4] S. Rana, P. White, M. Bradley, J. Comb. Chem. 2001, 3, 9 ± 15.
[5] a) K. Sundmacher, H. Künne, U. Kunz, Chem. Ing. Tech. 1998, 70,
267 ± 271; b) C. Altwicker, Dissertation, TU Clausthal, 2001.
[6] By employing monomers such as 4-vinylpyridine other polymers can
be incorporated into the porous glass rod by precipitation polymer-
ization.
[7] These studies were conducted as part of European-funded project at
the center for NMR spectroscopy of the University of Wageningen
(The Netherlands).
[8] Under these conditions, gel-type (Amberlite IRA 400) as well as
macroporous (Amberlite IRA 900) anion-exchange resins loaded
with borohydride anions, both of which are commercially available,
behave similarly to the bare monolithic rod and the powdered
material, respectively.
[9] Depending on the extent of polymer swelling caused by the solvent
1
used the pressure drop was 10 bar (3 mLmin
) and 28 bar
(10 mLmin 1), respectively, across a length of the reactor.
[10] G. Sourkouni-Argirusi, A. Kirschning, Org. Lett. 2000, 2, 3781 ± 3784.
[11] For automation common HPLC equipment can be used (e.g. pumps,
detectors, valves, dosing facilities).
[12] S. J. Haswell, R. J. Middleton, B. OꢁSullivan, V. Skelton, P. Watts, P.
Styring, Chem. Commun. 2001, 391 ± 398.
[13] The flow-through reactors presented are available from CHELO-
NA GmbH (Potsdam).
[*] Dr. W. Schrader
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1, 45470 Mülheim/Ruhr (Germany)
Fax : (‡49)208-306-2982
E-mail: wschrader@mpi-muelheim.mpg.de
Dr. J. Geiger
Institut für Spektrochemie und angewandte Spektroskopie (ISAS),
Institutsteil Berlin
Albert-Einstein-Strasse 9, 12489 Berlin-Adlershof (Germany)
Dr. M. Godejohann
Bruker Analytik GmbH
76287 Rheinstetten, Silberstreifen (Germany)
Dipl.-Chem. B. Warscheid, Dr. T. Hoffmann
Institut für Spektrochemie und angewandte Spektroskopie (ISAS),
Institutsteil Dortmund
Bunsen-Kirchhoff-Strasse 11, 44139 Dortmund (Germany)
[**] The financial support by the Bundesministerium für Bildung und
Forschung and the Senatsverwaltung für Wissenschaft, Forschung und
Kultur des Landes Berlin (ISAS Berlin), and the Ministerium für
Schule, Wissenschaft und Forschung NRW (ISAS Dortmund) is
gratefully acknowledged. The authors want to thank U. Marggraff and
W. Nigge (Dortmund) for assistence during MS measurements and Dr.
R. Köppe (Univ. Karlsruhe) and Dr. H. Sievers (GSG Karlsruhe) for
assistance in FT-ICR-MS measurements.
Supporting information for this article is available on the WWW under
http://www.angewandte.com or from the author.
3998
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4021-3998 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 21

COMMUNICATIONS
Scheme 1. Reaction pathway of a-pinene oxidation, which leads to the
intermediate Criegee biradicals 3.
For a better understanding of the natural aerosol formation
processes, ideally the complete set of products should be
known. Since synthetic standards are not readily available, the
most promising approach to reach this goal is the linking of
sophisticated analytical techniques, such as the coupling of
liquid chromatography (LC) to three different spectroscopic
methods, namely nuclear magnetic resonance (NMR) spec-
troscopy,[8] mass spectrometry (MS), and infrared (IR)
spectroscopy.[9]
LC/NMR/MS experiment are presented. This online experi-
ment is made possible through the combination of both a mass
spectrometer and NMR spectrometer with a gradient high-
pressure liquid chromatography (HPLC) system. The flow
coming from the LC is divided into two flows in the ratio 1:20.
The larger flow is introduced into a flow cell of the NMR
spectrometer, the smaller part is used to obtain corresponding
MS or MSn data. NMR scans can be obtained either in the on-
flow mode, which means a continuous flow through the NMR
The LC/NMR coupling, although not the most sensitive
method, clearly allows direct structural information about
compounds to be obtained. In Figures 1 and 2 the results of an
Figure 1. NMR spectra from cis-pinonic acid (4), obtained from an online experiment after LC separation; asignment of the signals using 1D, 2D NOESY
1
(nuclear overhauser enhancement spectroscopy) and 2D COSY (correlated spectroscopy). Integration of the 1D H NMR signal in combination with the
corresponding 2D COSY shows that signals Ha2 ± Ha4 and signals Hc1 and Hc2 (in total 5 protons) resonate at two positions around d ˆ 1.8 and 2.3 (the
correlation caused by J coupling between Ha1 and Ha2/Ha3 is indicated in the figure by arrows). The 2D NOESY spectrum indicates correlations caused by
dipole cross-relaxation between nuclei in a close spatial relationship. Some of the correlations not amenable to J-coupling spectroscopy are indicated.
Angew. Chem. Int. Ed. 2001, 40, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4021-3999 $ 17.50+.50/0
3999

COMMUNICATIONS
Figure 2. MS spectra from pinonic acid (4); [M] ˆ 184; [M(D1) D] ˆ 183; online LC/NMR/MS experiment. a) Positive ion MSn spectra (n ˆ 1 ± 5) and
b) Negative ion MSn spectra (n ˆ 1 ± 3), the proposed fragmentation is shown.
probehead while the spectrometer constantly aquires one-
dimensional (1D) proton spectra or in the loop-storage mode,
in which fractions are stored in capillaries (loops) during the
chromatography, the probes are then transfered to the NMR
and mass spectrometers. This set up allows MS5 spectra of 4 to
be obtained in the positive mode and even MS3 data in the
negative mode (Figure 2).
In addition to the important structural information about
reaction products, NMR spectroscopy allows the cis and trans
isomers of ozonolysis products from the a-pinene/ozone
reaction to be distinguished on the basis of the singlet
resonance signal of the methyl protons (Figure 3). The Hb1
The use of a FT-ICR-MS (Fourier-transform ion cyclotron
resonance mass spectrometer) allows not only high-resolution
mass data to be obtained, but also highly accurate mass
determination, which permits the calculation of the respective
molecular formula. Surprisingly, the high-resolution mass
spectra revealed a higher complexity of the products than
anticipated. While mass spectrometric observations with
standard resolution suggested only one compound at a certain
mass-to-charge ratio, the high-resolution mass spectra indi-
cate that more than one product with the same nominal mass
but different elemental composition can contribute to the
signal. In the case of a-pinene oxidation products, which can
only be composed of C, H, and O, the difference between CH4
(16.0313) and an oxygen atom (15.994914) can easily be
resolved by FT-ICR-MS. Figure 4 shows an example, were a
C10H15O4 ion (m/z 199.09769) is clearly separated from an
additional product with a composition of C9H11O5 (m/z
199.06133).
Figure 3. NMR spectra of cis-pinic acid (top) below trans-pinic acid (5)
(bottom), with in each case traces of other isomers.
and Hb2 signals of the methyl groups of cis-pinic acid (5) are
widely separated, while the difference between the corre-
sponding signals of the trans isomer is rather small. However,
in contrast to the results of recent field measurements,[10] only
cis isomers were found as reaction products when the reaction
was performed in the absence of light in the laboratory.
Figure 4. High-resolution mass spectrum of two reaction products at mass
m/z 199.
4000
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4021-4000 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 21

COMMUNICATIONS
molecular formula in square parentheses corresponds to negatively
charged ion).
Taking all of the spectrometric data into account a number
of new compounds (7, 8) can be identified with a high degree
of confidence. For all of the identified compounds, which
include the already described products (4 ± 6), new analytical
data could be obtained, that allows them to be included into a
reaction scheme of the a-pinene ozonolysis that is based
purely on identified compounds. In addition to these com-
pounds, a structural suggestion can be made for an additional
product 9 based on MS and IR data. This structure is
supported by the ester channel mechanism suggested by
Atkinson[11] to be one of the leading mechanisms for the
formation of terpene reaction products (Scheme 2).
cis-Pinonic acid (4): NMR: Ha1 2.990 (dd, 1H); Ha4, Hc1, Hc2 2.187 ± 2.348
(m, 3H); Hd 2.032 (s, 3H);[12] Ha2, Ha3 1.782 ± 1.905 (m, 2H); Hb1 1.266 (s,
3H); Hb2 0.797 (s, 3H); MS: 183.1027330 [C10H15O3] (0.33 ppm); IR: nÄ ˆ
1
2957, 1707, 1369, 1226, 1184, 948 cm
.
cis-Pinic acid (5): NMR: Ha1 2.769 (dd, 1H); Ha4, Hc1, Hc2 2.246 ± 2.401
(m, 3H); Ha3 2.018 ± 2.069 (m, 1H);[12] Ha2 1.794 (q, 1H); Hb1 1.167 (s, 3H);
Hb2 0.913 (s, 3H). MS: 185.0819860 [C9H13O4] (0.29 ppm); IR: nÄ ˆ 2961,
1
2657, 1713, 1419, 1246, 935, 740 cm
.
Hydroxy pinonic acid (6): NMR: Ha1 2.994 (dd, 1H); Ha4, Hc1, Hc2 2.182 ±
2.370 (m, 3H); Hd1, Hd2 4.093 ± 4.230 (dd, 2H); Ha2 1.891 (m, 1H); Ha3
possibly obscured; Hb1 1.204 (s, 3H); Hb2 0.787 (s, 3H); MS: 199.0976910
[C10H15O4] (0.54 ppm).
cis-nor-Pinonic acid (7): Ha1 3.100 (dd, 1H); Ha4 2.886 (dd, 1H); Ha2 2.352
(q, 1H); Hd 2.090 (s, 3H) (partly obscured by signal of acetonitrile); Ha3
1.852 (qd, 1H); Hb1 1.411 (s, 3H); Hb2 0.868 (s, 3H). MS: 169.0870480
[C9H13O3] (0.56 ppm).
cis-Pinornalic acid (8): NMR: Hd 9.631 (s, 1H); Ha1 2.817 (t, 1H); Hc1
2.575 (dd, 1H); Hc2 2.476 (dd, 1H); Ha4 2.408 (m, 1H); Ha2 possibly
obscured; Ha3 1.810 (dd, 1H); Hb1 1.193 (s, 3H); Hb2 0.917 (s, 3H). MS:
169.0870480 [C9H13O3] (0.56 ppm).
[Hydroxy acetic acid-((2,2-dimethyl)-cyclo-butyl)-ester] acetic acid (9):
MS: 215.0925990 [C10H15O5] (0.46 ppm); IR: nÄ ˆ 2976, 1742, 1392, 1275,
1
1130, 961 cm
.
Received: June 5, 2001 [Z17219]
[1] F. W. Went, Nature 1960, 187, 641.
[2] a) M. O. Andreae, P. J. Crutzen, Science 1997, 276, 1052; b) A. R.
Ravishankara, Science 1997, 276, 1058.
[3] a) T. Hoffmann, J. R. Odum, F. Bowman, D. Collins, D. Klockow,
R. C. Flagan, J. Seinfeld, J. Atmos. Chem. 1997, 26, 189; b) S. Koch, R.
Winterhalter, E. Uherek, A. Kolloff, P. Neeb, G. K. Moortgat, Atmos.
Environ. 2000, 34, 4031.
[4] a) R. Atkinson, S. M. Aschmann, J. Arey, B. Shorees, J. Geophys. Res.
1992, 97, 6065; b) B. R. Larsen, D. Di Bella, M. Glasius, R. Winter-
halter, N. R. Jensen, J. Hjorth, J. Atmos. Chem. 2001, 38, 231; c) M.
Hallquist, I. Wangberg, E. Ljungstrom, I. Barnes, K. H. Becker,
Environ. Sci. Technol. 1999, 33, 553.
Scheme 2. Proposed reaction scheme leading to the ozonolysis product 9
from 3b; in accordance to Atkinson.[11]
[5] a) R. Criegee, Angew. Chem. 1975, 87, 765; Angew. Chem. Int. Ed.
Engl. 1975, 14, 745; b) W. Adam, M. Braun, A. Griesbeck, V. Lucchini,
E. Staab, B. Will, J. Am. Chem. Soc. 1989, 111, 203; c) E. E. Royals,
L. L. Harrell, Jr., J. Am. Chem. Soc. 1955, 77, 3405.
[6] a) J. Yu, D. R. Crocker III, R. J. Griffin, R. C. Flagan, J. H. Seinfeld, J.
Atmos. Chem. 1999, 34, 207; b) S. Hatakeyama, K. Izumi, T.
Fukuyama, H. Akimoto, J. Geophys. Res. 1989, 94, 13013; c) Y.
Yokouchi, Y. Ambe, Atmos. Environ. 1985, 19, 1271; d) U. Kück-
elmann, B. Warscheid, T. Hoffmann, Anal. Chem. 2000, 72, 1905.
[7] a) M. E. Jenkin, D. E. Shallcross, J. N. Harvey, Atmos. Environ. 2000,
34, 2837; b) S. Koch, R. Winterhalter, A. Kolloff, E. Uherek, P. Neeb,
G. Moortgat in Proceedings of the EC/EUROTRAC-2 Joint Workshop
(Aachen, Germany) 1999, p. 300.
The results show that an analytical approach to the
investigation of atmospheric relevant organic aerosols by
coupling liquid chromatography with mass spectrometry,
NMR spectroscopy, and infrared spectroscopy allows valuable
information to be obtained about the structure of the reaction
products. Even with high-resolution mass spectrometry it is
not possible to differentiate between 7 and 8 which have the
same composition (empirical formula). But with the structural
information from both NMR and MS/MS a characterization is
possible.
With this instrumentation we have for the first time tools at
hand that are able to give enough information about such a
complex and difficult problem as terpene gas-phase oxidation.
While NMR spectroscopy allows cis and trans isomers to be
distinguished, in addition the structural information from
high-resolution and highly accurate mass determination make
it possible to address with precision the formulas of the
unknown products.
[8] a) G. Schlotterbeck, L.-H. Tseng, H. Händel, U. Braumann, K. Albert,
Anal. Chem. 1997, 69, 1421; b) M. Godejohann, M. Astratov, A. Preiû,
K. Levsen, C. Mügge, Anal. Chem. 1998, 70, 4104; c) U. G. Sidelmann,
D. Gavaghan, H. A. J. Carless, M. Spraul, M. Hofmann, J. C. Lindon,
I. D. Wilson, J. K. Nicholson, Anal. Chem. 1995, 67, 4441.
[9] J. Geiger, E. H. Korte, W. Schrader, J. Chromatogr. A 2001, 922, 99.
[10] I. G. Kavouras, N. Mihalopoulos, E. G. Stephanou, Nature 1998, 395,
683.
[11] R. Atkinson, J. Phys. Chem. Ref. Data 1997, 26, 215.
[12] Obscured by signal of acetonitrile, determination with selective total
correlation spectroscopy (TOCSY).
Experimental Section
Analytical data: NMR: (proton signals are referenced to the acetonitrile
(dˆ 1.93); MS: accurate mass determination (m/z: accuracy in parentheses;
Angew. Chem. Int. Ed. 2001, 40, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4021-4001 $ 17.50+.50/0
4001