Contrasting behavior of pentafluorophenoxyacetone and pentafluorobenzyloxyacetone in electron impact and electron capture mass spectrometry

Article abstract of DOI:10.1002/(SICI)1096-9888(199807)33:7<615::AID-JMS674>3.0.CO;2-N

Towards the goal of finding new ketone electrophores suitable as molecular labels for electrophoric release tags, pentafluorophenuxyacetone (1) and pentafluorbenzyloxyacetone (2) were prepared. Both ketones were evaluated by electron capture (EC) and electron impact (EI) modes of mass spectrometry (MS). By EC-MS, 1 nearly gave a single ion (as desired), whereas 2 gave many ions. This behavior was completely reversed in EI-MS. To account for certain ion fragments in the EC mass spectrum of 2, an anion radical MCLafferty-type rearrangement and loss of a carbene neutral were postulated. Electron impact of 1 gave an abundant ion at m/z 117 (C⁵F³⁺), which was suggested to be a diyne cation.

Full text of DOI:10.1002/(SICI)1096-9888(199807)33:7<615::AID-JMS674>3.0.CO;2-N

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 33, 615È620 (1998)
Contrasting Behavior of
PentaÑuorophenoxyacetone and
PentaÑuorobenzyloxyacetone in Electron Impact
and Electron Capture Mass Spectrometry
Linxiao Xu, Yulin Huang and Roger W. Giese*
Department of Pharmaceutical Sciences in the Bouve College of Pharmacy and Health Professions, Barnett Institute, and
Chemistry Department, Northeastern University, Boston, Massachusetts 02115, USA
Towards the goal of Ðnding new ketone electrophores suitable as molecular labels for electrophoric release tags,
pentaÑuorophenoxyacetone (1) and pentaÑuorbenzyloxyacetone (2) were prepared. Both ketones were evaluated by
electron capture (EC) and electron impact (EI) modes of mass spectrometry (MS). By EC-MS, 1 nearly gave a
single ion (as desired), whereas 2 gave many ions. This behavior was completely reversed in EI-MS. To account for
certain ion fragments in the EC mass spectrum of 2, an anion radical McLa†erty-type rearrangement and loss of a
carbene neutral were postulated. Electron impact of 1 gave an abundant ion at m/z 117 (C5F3‘), which was
suggested to be a diyne cation. ( 1998 John Wiley & Sons, Ltd.
KEYWORDS: electron capture mass spectrometry; electrophore; Fourier transform mass spectrometry; McLa†erty
rearrangement; carbene
brief summary of the subsequent literature appeared
more recently.4
INTRODUCTION
We are developing electrophores as covalent labels for
signal-based analytical methods such as DNA
sequencing1 and immunoassays.2 Ultimately the
electrophore-labeled reagent in the assay is measured by
chemically releasing the electrophore prior to its detec-
tion by electron capture mass spectrometry (EC-MS).
We are focusing on ketone electrophores since they can
be derived by oxidatively or thermally cleaving an
appropriate glycol linkage that connects the electro-
phore to the reagent of interest. The ideal ketone elec-
trophore for this purpose gives an abundant single ion
in the EC mass spectrum (aside from an associated
isotope peak). Since EC-MS can measure many di†erent
electrophores simultaneously, potentially many electro-
phore reagents can be combined to enhance the speed
and speciÐcity of signal-based assays.
Here we report the synthesis and testing of two struc-
turally similar ketone electrophores, each of which is an
acetone derivative. Whereas one ketone has the detec-
tion property we seek (it gives nearly a single abundant
ion by EC-MS), the other gives many fragmentation
products in the EC mass spectrum. Under electron
impact (EI) conditions, the fragmentation behavior of
the two ketone electrophores is reversed.
RESULTS AND DISCUSSION
In Fig. 1 is shown the synthesis and structures of the
two ketone electrophores that we studied by EC-MS
and EI-MS: pentaÑuorophenoxyacetone (1) and penta-
Ñuorobenzyloxyacetone (2). As seen, 1 was obtained by
converting acetone to 1-iodoacetone with iodine and
then reacting this intermediate product with pentaÑuo-
rophenol in the presence of potassium carbonate.
Similar reaction of 1-iodoacetone with pentaÑuoro-
benzyl alcohol failed to give 2, as did reaction of acetol
in reÑuxing acetone with pentaÑuorobenzyl bromide.
Synthesis of 2 was accomplished by reacting propargyl
alcohol with pentaÑuorobenzyl bromide, followed by an
acidic hydration reaction of the intermediate pentaÑuo-
robenzyl propargyl ether.
The fragmentation behavior of 1 and 2 in EC-MS
and EI-MS described here (Schemes 1È5) was observed
originally on a quadrupole EC/EI-MS instrument. The
compounds were later analyzed using an EC/EI-Fourier
transform (FT) MS system, which provided high-
resolution, accurate mass measurements.
Fewer ions could be seen with FTMS than with the
quadrupole instrument. For example, neither com-
pound gave a molecular ion in FTMS, under either EC
or EI conditions, and only some of the product ions
proposed in Scheme 2 could be seen. This is probably
because the ions are detected within a few milliseconds
in the quadrupole instrument, whereas the time is closer
to 1 s in FTMS. This provides more time for ions to
The analysis of Ñuorinated organic compounds by
mass spectrometry was reviewed many years ago3 and a
* R. W. Giese, Department of Pharmaceutical Sciences in the
Bouve College of Pharmacy and Health Professions, Barnett Institute,
and Chemistry Department, Northeastern University, Boston, Massa-
chusetts 02115, USA.
CCC 1076È5174/98/070615È06 $17.50
( 1998 John Wiley & Sons, Ltd.
Received 20 December 1997
Accepted 17 March 1998

616
L. XU, Y. HUANG AND R. W. GIESE
Figure 1. Synthesis of 1 and 2.
fragment or rearrange in the FTMS instrument. For the
ions that could be observed in FTMS, all of the accu-
rate mass measurements were consistent with the ele-
mental compositions that were expected (see Table 1).
Attempts to obtain stepwise fragmentation informa-
tion by MS/MS experiments with the FTMS instrument
were not even partly successful, however. No fragment
ions were observed when the parent ion from EC
(highest mass ion observed for each compound) was
subjected to collisionally induced dissociation (CID).
Perhaps this was a consequence, to at least some degree,
of electron stripping by CID.5 In the EI mode, numer-
ous ions were generated from CID on the parent ion,
precluding deÐnitive evaluation of the proposed frag-
mentation pathways.
from the parent anion radical. Additional loss of a Ñuo-
rine atom explains the minor peak at m/z 164.
In contrast, many fragmentation products form when
2 is subjected to EC-MS, as seen in Fig. 2(B). Suggested
structures and fragmentation pathways for the major
ions are presented in Scheme 1.
A simple loss of a Ñuorine atom accounts for m/z 235,
which could, in turn, lose a hydrogen atom, driven by a
ring closure, to form m/z 234. The latter species also
might arise from a single-step elimination of HF. Loss
of HF has been reported by others in the analysis of
polyÑuoroaromatic compounds by EC-MS.6h8 The
fragment ion at m/z 215 probably originates from the
radical anion [C H O F ]~~ at m/z 234 by the loss of
10 6 2 4
another Ñuorine atom. The subsequent transition of m/z
As seen in Fig. 2(A), 1 gives nearly a single ion by
EC-MS, corresponding to a pentaÑuorophenoxy anion
(m/z 183) arising from the loss of an acetone radical
234 to m/z 214 is the same type of ring closure reaction,
including the possibility of two pathways, analogous to
the conversion of m/z 254 to 234. A simple loss of an
acetyl neutral, perhaps involving an electron transfer
from the aromatic nucleus to the benzylic carbon of m/z
234, explains m/z 191. The proposed ring closure pro-
ducts at m/z 162 and m/z 178 also appear to derive from
m/z 235. Replacement of an F by OH and O probably
explains m/z 252 and 251, respectively.
Table 1. Elemental composition measurement of compounds 1
and 2 by FTMS
Calculated mass
(u)
Experimental
Compound
Mode
Formula
accuracy (ppm)
The formation of the fragment with m/z 179 is inter-
esting. We suggest that it arises from m/z 235 starting
with a proton migration followed by loss of an acetyl
carbene. This postulated mechanism is shown in more
detail in Scheme 2. It is well known that carbanions
connected directly to leaving groups, which are named
carbenoids, can generate carbenes by an a-elimination
process.9 The released acetylcarbene should Ðnally rear-
range to a corresponding methylketene via the Wol†
rearrangement.10 Also shown in Scheme 2 is a fragmen-
tation pathway from the enolate form of m/z 235 to
account for m/z 72 (oxyenolate anion radical) and m/z
71 (oxyketene anion derived from m/z 72 by loss of a
hydrogen atom). This pathway also is summarized in
Scheme 1.
We suggest that a McLa†erty-type rearrangement is
mainly responsible for the formation of the apparent
pentaÑuorobenzaldehyde anion radical at m/z 196 (see
Scheme 1), as illustrated in more detail in Scheme 3. An
anionic McLa†erty-type rearrangement has been sug-
gested before.13 The fragment at m/z 196 might also
1
EC
EI
C F O
6 5
C H
182.9874
91.05422
116.9947
154.9915
166.9915
181.0071
182.9864
197.0020
240.0112
3.0
1.1
1.3
1.5
2.5
1.9
3.0
1.6
1.8
7
7
C F
5 3
C F
5 5
C F
6 5
C F H
7 5
2
C F O
6 5
C F H O
7 5
C F H O
9 5
5
2
5
2
2
EC
C F
6 4
C F
6 5
C F H O
7 4
C F H O
147.9941
166.9926
178.0047
191.0125
195.9953
210.9824
214.0247
91.05422
149.0209
161.0009
181.0071
194.9864
2.1
4.2
2.6
3.5
1.8
1.5
1.2
1.2
3.3
1.7
1.7
1.6
2
8 4
3
C F HO
7 5
C F O
7 5
2
C
F H O
10 3
5 2
EI
C H
7
7
C F H O
6 3
C F H
7 4
4
C F H
7 5
2
C F O
7 5
( 1998 John Wiley & Sons, Ltd.
J. Mass Spectrom. 33, 615È620 (1998)

EI AND EC MASS SPECTROMETRY OF ACETONE DERIVATIVES
617
Figure 2. Mass chromatograms of 1 (M ¼239 Da) by EC-MS and EI-MS (A and C, respectively) and similarly of 2 (M ¼254 Da) (B and
D, respectively).
r
r
arise by loss of a hydrogen atom from m/z 197, as
shown in Scheme 1.
leads to a major ion at m/z 155 from m/z 183 that forms
only to a minor degree from m/z 184, as indicated in
Scheme 5. However, once the m/z 155 ion losses a Ñuo-
rine atom, forming m/z 136, a common pathway takes
place leading to m/z 117. Not shown, but conceivable, is
that m/z 155 might yield m/z 117 without proceeding
through m/z 136. The signiÐcant intensity of the m/z 117
peak implies a relatively stable species such as the diyne
cation shown. Previously a structure for this ion has not
been proposed.
In contrast to its EC mass spectrum, the EI mass
spectrum of 2, shown in Fig. 2(D), is nearly a single
peak, at m/z 181, which is a pentaÑuorobenzyl or penta-
Ñuorotropylium cation. The weak peak at m/z 161
apparently derives from m/z 181 by loss of HF. No
molecular ion (m/z 254) is observed. (However, the posi-
tive chemical ionization mass spectrum of 2 displays an
[M ] H]` peak at m/z 255; data not shown.)
Whereas 1 gives a simple EC mass spectrum, its EI
mass spectrum is complex, as seen in Fig. 2(C), which is
partly rationalized in Scheme 4. The molecular cation
radical gives the most intense peak (m/z 240). Single-
bond cleavage reactions can explain the fragments at
m/z 197, 183 and 167. The fragment at m/z 167 may also
EXPERIMENTAL
Reagents
come from the loss of CH O from the cation at m/z 197
2
(see Scheme 4), since Russell et al.14 observed a loss of
CH O from the analogous C H OCH `ion derived
PentaÑuorophenol (99]%), pentaÑuorobenzyl bromide
(99]%), propargyl alcohol (99%), triethylbenzyl-
ammonium chloride (TEBA), 18-crown-ether and other
reagents were purchased from Aldrich Chemical
(Milwaukee, WI, USA). All solvents used were obtained
from Doe & Ingalls Chemicals (Medford, MA, USA).
2
6 5
2
from non-Ñuorinated phenoxyacetone.
Others have observed three major ions when penta-
Ñuorphenol is subjected to EI-MS; m/z 184
([C F OH]`~, the M`~), 136 ([C F ]`~ and 117
6 5
5 4
([C F ]`).15 Of the minor ions, one appeared to be at
5 3
m/z 155 (our interpretation of the mass spectrum shown
in Ref. 15). Interestingly, we also observe major ions at
Instrumentation
m/z 136 and 117, along with an ion fragment at m/z 183
([C F O]`), plus a major ion at m/z 155. We believe
6 5
that this all can be accounted for by the fragmentation
Gas chromatography/electron capture mass spectrom-
etry was performed on a Hewlett-Packard Model
5988A mass spectrometer connected to a Hewlett-
Packard Model 5890 Series II gas chromatograph. The
pathways shown in Scheme 5. Apparently the quantitat-
ive behavior of the m/z 183 and 184 ions initially is dif-
ferent because only the latter can discharge HF. This
( 1998 John Wiley & Sons, Ltd.
J. Mass Spectrom. 33, 615È620 (1998)

618
L. XU, Y. HUANG AND R. W. GIESE
Scheme 1
Scheme 2
Scheme 3
( 1998 John Wiley & Sons, Ltd.
J. Mass Spectrom. 33, 615È620 (1998)

EI AND EC MASS SPECTROMETRY OF ACETONE DERIVATIVES
619
Scheme 4
ion source temperature was 250 ¡C; the electron energy
was 240 eV and the methane ion source pressure was 2
Torr (1 Torr \ 133.3 Pa). Helium was the carrier gas for
a 25 m ] 0.11 mm ] 0.17 lm Hewlett-Packard Ultra 1
capillary column. Gas chromatography/electron impact
(70 eV) mass spectrometry was performed on a Finni-
gan Model 4021B quadrupole mass spectrometer inter-
obtain stepwise fragmentation information by MS/MS.
Ions of interest were isolated by a sweep pulse which
ejected all other ions coexisting inside the ICR ion trap.
Nitrogen was introduced into the analyzer by a pulsed
valve as collision gas. A resonant excitation pulse
(18 ks) was used to achieve CID. No fragment ions were
observable for parent ions generated from EC. For the
parent ions generated from EI, CID generated numer-
ous fragment ions in the m/z region between 80 and 130.
faced to
a
Hewlett-Packard Model 5890A gas
chromatograph. 1H and 13C NMR spectra were
obtained on a Varian XL-300FT instrument with
CDCl as the solvent and tetramethylsilane as the inter-
nal standard. Preparative gas chromatography was
3
PentaÑuorophenoxyacetone (1)
carried out on a Varian 3300 instrument with helium as
the carrier gas using a 20% SE-30/Chromosorb W
(non-acid washed, 60È80 mesh), 10@ ] 1/4A aluminum
column.
Iodine (2.54 g (10 mmol)) was added to 15 ml of acetone
followed by stirring for 2 h at room temperature. The
reaction mixture was added dropwise into a stirred
mixture of 0.92 g (5 mmol) of pentaÑuorophenol, 5 g of
potassium carbonate, 0.1 g of triethylbenzylammonium
chloride and 30 ml of acetone. After reÑuxing for 3 h,
the inorganic substances were removed by Ðltration.
The solvent was evaporated, the residue was dissolved
in 30 ml of pentane and 3 g of silica gel were added for
decolorization. Preparative gas chromatography gave
0.28 g (23%) of the product as a colorless liquid. 1H
NMR, d (ppm) 2.29 (s, 3H, CH ), 4.74 (s, 2H, CH ); 13C
Elemental compositions of the ions observed in
EI-MS and EC-MS were determined by making high-
resolution, accurate mass measurements with a Bruker
APEX 47e FTMS instrument (Bruker Instruments, Bill-
erica, MA, USA) equipped with an inÐnity cylindrical
ion trap and a 4.7 T superconducting magnet. Samples
were introduced through a heated external direct probe
insertion port located on the side of the source vacuum
chamber. The ion source was kept at 200 ¡C and the
electron energy was 70 eV for both EI and EC. Ions
generated at the source were transported to the ICR
trap through an electrostatic ion guide. The instrument
was calibrated by a two-parameter least-squares Ðtting
procedure using the same experimental conditions.
Three ions generated from known ketones synthesized
in our laboratory were used as calibration standards at
m/z 135.0451, 163.0764 and 195.0663, respectively. The
mass accuracy was in the low ppm range.
3
2
NMR, d (ppm) 26.1 (CH ), 77.6 (CH , 4J
\ 3 Hz),
3
2
C,F
137.9 (C-2@ and C-6@, J \ 250 Hz), 139.0 (C-4@, J
\
C,F
C,F
248 Hz), 141.1 (C-3@ and C-5@, J \ 245Hz), 169.5 (C-
C,F
1@), 202.8 (C \ 0).
PentaÑuorophenyl propargyl ether (3)
A mixture of 1.3 g (5 mmol) of pentaÑuorobenzyl
bromide, 0.31 g (5.5 mmol) of propargyl alcohol, 50 mg
CID was also explored in an unsuccessful attempt to
Scheme 5
( 1998 John Wiley & Sons, Ltd.
J. Mass Spectrom. 33, 615È620 (1998)

620
L. XU, Y. HUANG AND R. W. GIESE
of 18-crown and 5 g of potassium carbonate in 20 ml of
acetone was stirred under reÑux for 10 h. After the
removal of the inorganic substances by Ðltration, the Ðl-
trate was evaporated to remove most of the solvent.
Preparative gas chromatography a†orded 0.9 g (76%) of
product as a colorless liquid. 1H NMR, d (ppm) 2.51 (t,
of diethyl ether were added. The water layer was
separated and the organic layer was dried over anhy-
drous magnesium sulfate, giving, after solvent removal,
0.83 g (93%) of product as white crystals, m.p. 54È55 ¡C.
1H NMR, d (ppm) 2.18 (s, 3H, CH ), 4.18 (s, 2H, CH ),
3
2
4.74 (t, 2H, 4J \ 1.8 Hz, CH Ar); 13C NMR, d (ppm)
H,F
2
1H, 4J \ 2.4 Hz, HwCy), 4.22 (d, 2H, 4J \ 2.4 Hz,
25.4 (CH ), 59.5 (CH Ar), 75.2 (CH ), 110.4 (C-1@,
H,F
3
2
2
CH ), 4.70 (t, 2H, 4J \ 1.8 Hz, CH Ar); 13C NMR, d
2J \ 18 Hz), 137.2 (C-2@ and C-6@, J \ 253 Hz),
2
H,F
2
C,F
C,F
(ppm) 57.9 (CH ), 58.3 (CH Ar), 75.2 (HwCy, J
\
141.2 (C-4@, J \ 255 Hz), 145.4 (C-3@ and C-5@, J
\
2
2
C,H
C,F
C,F
251 Hz), 78.5 (wCy), 110.8 (C-1@, 2J \ 19 Hz), 137.5
253 Hz), 205.1 (C \ 0); MS (PCI), m/z (%)
255([M ] H]`, 0.5), 181 (100), 161 (9), 58 (25), 44 (22),
43 (58).
C,F
(C-2@ and C-6@, J \ 253 Hz), 141.5 (C-4@, J \ 255
C,F
C,F
Hz), 145.8 (C-3@ and C-5@, J \ 250 Hz).
C,F
PentaÑuorobenzyloxyacetone (2)
Acknowledgements
To a stirred mixture of 0.1 g of sulfuric acid and 1 g of
mercury(II) sulfate in 50 ml of methanolÈwater (3 :1),
0.83 g (3.5 mmol) of 3 was added dropwise slowly at
room temperature followed by stirring under reÑux for
2 h. After the evaporation of most of the solvent, 20 ml
This work was supported by NIST Award 70NANB5H1038 and NIH
Grant HG00029. We thank James L. and Faith P. Waters for their
generous support in establishing the FTMS system at the Barnett
Institute. Susan Wolf and Manasi Saha provided technical assistance.
This paper is Contribution No. 737 from the Barnett Institute.
REFERENCES
1. L. Xu, N. Bian, Z. Wang, S. Abdel-Baky, S. Pillai, D. Magiera,
V. Murugaiah, R. W. Giese, P. Wang, T. O’Keeffe, H. Abusha-
maa, L. Kutney, G. Church, S. Carson, D. Smith, M. Park, J.
Wronka and F. Laukien, Anal. Chem. 69, 3595 (1997).
2. N. Bian, P. Wang, Z. Wang, L. Xu, G. Church and R. W. Giese,
Rapid Commun. Mass Spectrom. 11, 1781 (1997).
3. J. R. Majer, in Advances in Fluorine Chemistry, ed. by M.
Stacey, J. C. Tatlow and A. G. Sharpe, Vol. 2, p. 55. Butter-
worths, London (1961).
4. D. T. Meshri, in Chemistry of Organic Fluorine Compounds II,
ed. by M. Hudlicky and A. E. Pavlath, ACS Monograph No.
187, p. 1031. American Chemical Society, Washington, DC
(1995).
8. D. A. Durden, B. A. Davis and A. A. Boulton, Biol. Mass
Spectrom. 20, 375 (1991).
9. W. Kirmse, Carbene Chemistry, 2nd edn. Academic Press,
New York (1971).
10. K. P. Zeller, Liebigs Ann. Chem. 2036 (1979).
11. M. Zollinger and J. Seibl, Org. Mass Spectrom. 20, 649
(1985).
12. U. P. Schlunegger, Advanced Mass Spectrometry, Applica-
tions in Organic and Analytical Chemistry, p. 77. Pergamon
Press, New York (1980).
13. C. K.-C. Low and A. M. Duffield, Biomed. Mass Spectrom.
12, 348 (1985).
14. D. H. Russell, B. S. Freiser, E. H. McBay and D. C. Canada,
Org. Mass. Spectrom. 18, 474 (1983).
15. L. D. Smithson, A. K. Bhattacharya and C. Tamborski, Org.
Mass Spectrom. 4, 1 (1970).
5. W. B. Knighton, L. J. Sears and E. P. Grimsrud, Mass
Spectrom. Rev. 14, 327 (1995).
6. T. M. Trainor and P. Vouros, Anal. Chem. 59, 601 (1987).
7. S. Tajima, M. Ueki, S. Tajima, O. Sekiguchi and A. Shigihara,
Rapid Commun. Mass Spectrom. 10, 1076 (1996).
( 1998 John Wiley & Sons, Ltd.
J. Mass Spectrom. 33, 615È620 (1998)