Chemistry of bifunctional photoprobes. 1. Perfluoroaryl azido functionalized phosphorus hydrazides as novel photoreactive heterobifunctional chelating agents: High efficiency nitrene insertion on model solvents and proteins

Article abstract of DOI:10.1021/jo961867b

Synthesis and evaluation of a new class of photochemically activated heterobifunctional chelating agents for protein modification is described. Selective functionalization of perfluoroaryl azides by versatile phosphorus hydrazide ligating systems 2 and 3 for the complexation of transition metals and analogous radiometals form the basis for these new agents. The utility of the photogenerated precursors from these bifunctional agents to form covalent attachments is demonstrated through examination of C-H bond insertion on cyclohexane. Representative amide-coupled phosphorus hydrazides 5 and 6 provide > 78% insertion of the probe into unactivated C-H bonds of cyclohexane with short photolysis times. Photoconjugation of the photoactivable heterobifunctional chelating agent 6 and its Pd metalated analog 7 with HSA is also evaluated. The uncomplexed chelate appears to add to HSA with high efficiency, consistent with the observed 82% bond insertion into model solvents. Covalent attachment of 7, evaluated through the use of ¹⁰⁹Pd, was estimated to be between 49% and 74% with the uncertainty arising because of prephotolysis association of the ¹⁰⁹Pd complex with HSA. The application of in situ ¹⁹F NMR to distinguish between bond insertion and noninsertion processes is demonstrated. These results suggest that functionalized perfluoroaryl azido phosphorus hydrazides may find utility as heterobifunctional photolabeling agents for attaching radionuclides to proteins and antibodies.

Full text of DOI:10.1021/jo961867b

2
798
J . Org. Chem. 1997, 62, 2798-2807
Ch em istr y of Bifu n ction a l P h otop r obes. 1. P er flu or oa r yl Azid o
F u n ction a lized P h osp h or u s Hyd r a zid es a s Novel P h otor ea ctive
Heter obifu n ction a l Ch ela tin g Agen ts: High Efficien cy Nitr en e
In ser tion on Mod el Solven ts a n d P r otein s
†
†
§
,†
Raghoottama S. Pandurangi, Srinivasa R. Karra, Kattesh V. Katti, Robert R. Kuntz,* and
‡
Wynn A. Volkert
Department of Chemistry, University of Missouri, Columbia, Missouri 65211, H. S. Truman VA Hospital,
Columbia, Missouri 65211, and Center for Radiological Research and MU Research Reactor, Allton
Buildings Laboratories, 301, Business Loop 70W, Columbia, Missouri 65203
Received October 2, 1996X
Synthesis and evaluation of a new class of photochemically activated heterobifunctional chelating
agents for protein modification is described. Selective functionalization of perfluoroaryl azides by
versatile phosphorus hydrazide ligating systems 2 and 3 for the complexation of transition metals
and analogous radiometals form the basis for these new agents. The utility of the photogenerated
precursors from these bifunctional agents to form covalent attachments is demonstrated through
examination of C-H bond insertion on cyclohexane. Representative amide-coupled phosphorus
hydrazides 5 and 6 provide >78% insertion of the probe into unactivated C-H bonds of cyclohexane
with short photolysis times. Photoconjugation of the photoactivable heterobifunctional chelating
agent 6 and its Pd metalated analog 7 with HSA is also evaluated. The uncomplexed chelate appears
to add to HSA with high efficiency, consistent with the observed 82% bond insertion into model
solvents. Covalent attachment of 7, evaluated through the use of 1 Pd, was estimated to be between
09
1
09
4
9% and 74% with the uncertainty arising because of prephotolysis association of the Pd complex
1
9
with HSA. The application of in situ F NMR to distinguish between bond insertion and
noninsertion processes is demonstrated. These results suggest that functionalized perfluoroaryl
azido phosphorus hydrazides may find utility as heterobifunctional photolabeling agents for
attaching radionuclides to proteins and antibodies.
In tr od u ction
The development of tumor-specific radiopharmaceuti-
We are currently investigating the efficacy of radiola-
beling of biomolecules via photochemical methods using
bifunctional labels containing a photoprobe and a che-
cals continues to be an important area of research
because of their applications in the diagnosis and therapy
lated radionuclide at the termini. Aromatic azides have
10,11
been widely used for photoaffinity labeling,
but
1
-3
of human cancer.
Attachment of radiolabels to bio-
unproductive side reactions competing with the formation
and insertion reactions of reactive singlet nitrene inter-
mediates lead to low efficiencies. To achieve the high
yields and optimum in vivo stability required for radio-
labeling, it is crucial that the photoprobe exhibit a very
high efficiency for covalent binding to the biomolecule.
A high efficiency of conjugation also facilitates the
molecules provides a viable approach for detection and
subsequent therapy of certain cancers.4 Conventional
methods for radiolabeling of biomolecules require the
intermediacy of two or more synthetic steps and the use
of substrates which may be antagonistic to the biomol-
ecule. These factors have frequently contributed to low
radiolabeling yields and denaturation of the biomol-
ecules,6 suggesting the need for development of alterna-
tive techniques to conventional chemical affinity label-
ing.8
,5
1
2
isolation of labeled radioimmunoconjugates in pure form.
,7
(8) For reviews see: (a) Bruner, J . Annu. Rev. Biochem. 1993, 62,
,9
483. (b) Meares, C. F.; Wensel, T. G. Acc. Chem. Res. 1984, 17, 202. (c)
Brinkley, M. Bioconjugate Chem. 1992, 3, 2.
(
9) (a) Iankowski, K. J .; Parker, D. Diagnosis and Therapy with
*
Author to whom correspondence should be addressed.
University of Missouri.
H. S. Truman VA Hospital.
Center for Radiological Research and MU Research Reactor.
Abstract published in Advance ACS Abstracts, April 15, 1997.
Antibody Conjugates of Metal Radioisotopes. In Advances in Metals
in Medicine; Abrams, M. J ., Murrer, B. A., Eds.; J AI Press, Inc: New
York, 1993; Vol. 1, pp 29-73 and references therein. (b) Parker, D.
Chem. Soc. Rev. 1980, 19, 271. (c) Meares, C. F.; Moi, M. K.; Diril, H.;
Kukis, D. L.; McCall, M. J .; Deshpande, S. V.; Denardo, S. T.; Snook,
D.; Epenatos, A. A. Br. J . Cancer (suppl.) 1990, 10, 21. (d) Gamlow, O.
A. Nucl. Med. Biol. 1991, 18, 369. (e) Franz, J .; Volkert, W. A.;
Barefield, E. K.; Holmes, R. A. Nucl. Med. Biol. 1987, 14, 569. (f)
Abrams, M. J .; J uweid, M.; Tenkate, C. I.; Schwartz, D. A.; Hauser,
M. M.; Gaul, F. E.; Fucello, A. J .; Rubin, R. H.; Strauss, H. W.;
Fischman, A. J . J . Nucl. Med. 1990, 31, 2022. (g) Fritzberg, A. R.;
Kasina, S.; Vanderheyden, J . L.; Srinivasan, A. Eur. Pat. Appl. EP
284071, 1988.
†
‡
§
X
(
1) Stadalink, R. C.; Vera, D. R.; Krohn, K. A. Receptor-Binding
Radiopharmaceuticals: Experimental and Clinical Aspects. In Nuclear
Medicine Annual; Leonard, M., Weissmann, H. S., Eds.; Raven Press:
New York, 1986; pp 105-139.
(
2) Dewanjee, M. K. Semin. Nucl. Med. 1990, 20, 5.
(3) Deutsch, E.; Libson, K.; J urisson, S. Prog. Inorg. Chem. 1983,
3
0, 75.
4) Keenan, A. M. Radiolabeled Monoclonal Antibodies: Current
(
Status and Future Outlook. In Nuclear Medicine Annual; Leonard, M.,
Weissmann, H. S., Eds.; Raven Press: New York, 1988; pp 171-205.
(10) (a) Schuster, G. B.; Platz, M. S. Advances in Photochemistry;
Volman, D., Hammond, G., Neckers, D., Eds.; Wiley: New York, 1991;
Vol. 17, p 69. (b) Scriven, E. F. Y. In Reactive Intermediates; Abramo-
vitch, R. A., Ed.; Plenum: New York, 1981; Vol. 2, Chapt. 1. (c)
Lwowski, W. Reactive Intermediates; J ones, M., Moss, R. A., Eds.;
Wiley: New York, 1981; Chapter 8. (d) Reiser, A.; Wagner, H. M. The
Chemistry of the Azido Group; Patai, S., Ed.; Wiley: New York, 1971.
(e) Shrock, A. K.; Schuster, G. B. J . Am. Chem. Soc. 1984, 66, 5228.
(
5) Buchsbaum, D. J .; Lawrence, T. S. Antibody, Immunoconjugates
Radiopharm. 1991, 4, 245.
6) Colcher, D.; Keenan, A. M.; Larson, S. M.; Schlom, J . Cancer
Res. 1984, 44, 5744.
7) Min, L.; Meares, C. F.; Zhong, G. R.; Miers, L.; Xiong, C. Y.; De
Nardo, S. J . Bioconjugate Chem. 1994, 5, 101.
(
(
S0022-3263(96)01867-1 CCC: $14.00 © 1997 American Chemical Society

Chemistry of Bifunctional Photoprobes
J . Org. Chem., Vol. 62, No. 9, 1997 2799
The groups of Platz13 and Keana14 have pioneered the
development of polyfluorinated aromatic azides as singlet
nitrene precursors which minimize the competitive path-
ways and enhance the insertion efficiency. Subsequently,
we demonstrated that 4-azidotetrafluorobenzonitrile,
upon photolysis, gave the highest yield of insertion into
unactivated C-H bonds yet reported.15 The photochemi-
cal attachment of polyfluorinated aryl azides to proteins
is also characterized by high insertion efficiencies16 and
retention of immunoreactivity (>99%) in the postlabeling
stage.17 In fact, for a model photoprobe, the extent of
bond insertion into proteins (HSA, IgG) exceeds the C-H
insertion efficiency in cyclohexane.16 This procedure does
not involve external chemical activating agents, therefore
minimizing physical and chemical manipulations of
labeled biomolecules. While tritiated aryl azides have
been extensively used to study the binding site of
receptors, the functionalization of ligands with photola-
bels and their subsequent use in radiolabeling studies
nuclide, 1.02 MeV, σ ) 12 barn, medium range â source
1
9
for therapy ) make it an ideal candidate for use in
conjunction with chelating agents for potential radioim-
munoconjugates.20
Herein we describe the synthesis and chemistry of the
photoprobe-ligand complex utilizing a perfluoroaryl
azide photoprobe and a versatile phosphorus hydrazide
ligating system for the complexation of radiometals. The
light-induced insertion of the photoprobe-ligand com-
plexes into unactivated C-H bonds in cyclohexane is
demonstrated by a quantitative comparison of the pho-
toadduct with the independently synthesized cyclohexane
adduct. The photoconjugation of this probe with the
model protein human serum albumin (HSA) with and
without chelation to 1 Pd is used to evaluate the poten-
tial of this technique for the labeling of biomolecules. A
convenient screening technique, based on 19F NMR, for
evaluation of bond insertion reactions of the modified
photoprobes is also described. This report forms the first
quantitative analysis of the efficiency of photoconjugation
by a radioanalytical method corroborated by spectroscopy
09
are mostly limited to the pioneering work of Katzenel-
lenbogen et al.18 The next logical step is establishment
(
UV/vis and 19F NMR). The current system is an example
of a link between the photolabel and the ligating moiety.
The photolabel-functionalized ligand can then be labeled
for future generations of heterobifunctional photoprobe
chelating systems carrying diagnostic and therapeutic
radionuclei.
with γ- or â-emitting radionuclides (99mTc,
109
Pd) for
diagnostic or therapeutic applications, respectively. The
favorable therapeutic values of 1 Pd (â-emitting radio-
09
Resu lts a n d Discu ssion
(
11) For photoaffinity labeling: (a) Chavan, A. J .; Richardson, S.
K.; Kim, H.; Haley, E. B.; Watt, D. S. Bioconjugate Chem. 1993, 4,
68. (b) Birnbaumer, M.; Schrader, W. T.; O’Malley, W. B. J . Biol.
In tr od u ction of P er flu or oa r yl Azid o Moieties on
P h osp h or u s Hyd r a zid es. The choice of phosphorus
hydrazides as the chelating moiety is based on the
following characteristics: (a) these compounds are mul-
tifunctional so that the incorporation of the photolabel
still leaves ligating moieties for chelating metallic radio-
isotopes, (b) phosphorus hydrazides are inert under
photolytic conditions, allowing only the attached photo-
label to undergo photolytic transformations, and (c)
radiolabeling with metallic radioisotopes is versatile
because phosphorus hydrazides offer a combination of
electron donor centers in the form of hard-nitrogen bases
and phosphorus chalcogenides.21
2
Chem. 1983, 255, 1637. (c) Kerwin, B. A.; Yount, R. G. Bioconjugate
Chem. 1992, 3, 328. (d) Rogers, G. A.; Parsons, S. M. Biochemistry 1992,
1, 5770. (e) Ibarrola, I.; Alejandro, A.; Marino, A.; Sancho, M. J .;
Marcarulla, J . M.; Trueba M. J . Membrane Biol. 1992, 125, 185. (f)
Galardy, R. E.; Craig, L. C.; J amieson, J . D.; Printz, M. P. J . Biol.
Chem. 1974, 249(11), 3510. (g) Cavalla, D.; Neff, N. H. Biochem.
Pharmacol. 1985, 34, 1821. (h) Haley, B. E. Methods Enzymol. 1991,
00, 477. (i) Photochemical Probes in Biochemistry; Neilson, P. E., Ed.;
NATO ASI Series C, Vol. 272; Kluwer Academic Publishers: Boston,
989. (j) Lewis, C. T.; Haley, B. E.; Carlson, G. M. Biochemistry 1989,
8, 9248. (k) Turro, N. J . Modern Molecular Photochemistry;
Benjamin/Cummings: New York, 1978. (l) Bayley, H.; Staros J . W.
Photoaffinity labeling and Related Techniques. In Azides and Nitrenes,
Reactivity and Techniques; Scriver, E. F. V., Ed.; Academic Press: New
York, 1984; Chap. 9, p 433. (m) Bayley, H. Reagents for Photoaffinity
Labeling. In Photogenerated Reagents in Biochemistry and Molecular
Biology; Work, T. S., Burdon, R. H., Eds.; Elsevier: Amsterdam, 1983;
Chap. 3, p 26.
3
2
1
2
The bis(hydrazido)phosphine sulfide 1 is highly reac-
tive toward nucleophilic substitution reactions at the NH
centers. In order to selectively direct a photolabel to a
2
(12) (a) Cavalia, D.; Neff, N. H. Biochem. Pharmacol. 1985, 34, 2821.
single NH
other NH
2
group of 1, it was necessary to protect the
as described in Scheme 1. The reaction of
(
2
b) Choudary, V.; Westheimer, F. H. Annu. Rev. Biochem. 1979, 48,
93.
2
(13) (a) Leyva, E.; Young, M. J . T.; Platz, M. S. J . Am. Chem. Soc.
m-nitrobenzaldehyde with 1 in 1:1 stoichiometry at 0 °C
is clean and highly selective and gave the mono Schiff
1
1
986, 108, 8307. (b) Young, M. J . T.; Platz, M. S. Tetrahedron Lett.
989, 30, 2199. (c) Poe, R.; Grayzar, J .; Young, M. J . T.; Leyva, E.;
Schnapp, K. A.; Platz., M. S. J . Am. Chem. Soc. 1991, 113, 3209. (d)
Poe, R.; Schnapp, K.; Young, M. J . T.; Grayzar, J .; Platz., M. S. J . Am.
Chem. Soc. 1992, 114, 5054. (e) Soundararajan, M.; Platz, M. S. J .
Org. Chem. 1990, 55, 2034. (f) Young, M. J . T.; Platz, M. S. J . Org.
Chem. 1991, 56, 6403. (g) Levya, E.; Munoz, D.; Platz, M. S. J . Org.
Chem. 1989, 54, 5938.
(19) Ehrhardt, G. J .; Ketring, A. R.; Volkert, W. A. Production of
Isotopes at Nuclear Reactors. In Synthesis and Applications of Isoto-
pically Labeled Compounds; Buncel, E., Kabalka, G. W., Eds.; Elsevier
Science: The Netherlands, 1991; pp 159-164.
(20) (a)Kopper, G. A. Bioconjugate Chem. 1990, 1, 13. (b) Brinkley,
M. Bioconjugate Chem. 1992, 3, 2. (c) Waldemann, T. A. Science 1991,
252, 1657. (d) Hnatowich, D. J . Recent developments in the radiola-
beling of antibodies with iodine, indium, and technetium. In Seminars
in Nuclear Medicine; Freeman, L. M., Blaufox, M. D., Eds.; W. B.
Saunders: Philadelphia, 1990; Vol. XX, pp 80-91. (e) DeLand, F. H.;
Goldenberg, D. M. Radiolabeled Antibodies: Radiochemisry and Clini-
cal Applications. In Clinical Radionuclide Imaging; Freeman, L. M.,
Ed.; Grune and Stratton: Orlando, 1986; Vol. 3, pp 1915-1992.
(21) (a) Katti, K. V.; Singh, P. R.; Barnes, C. L. Inorg. Chem. 1992,
31, 4588. (b) Majoral, J . P.; Kramer, R.; Nvech, J .; Mathis, F.
Tetrahedron 1976, 32, 2633. (b) Katti, K. V.; Singh, P. R.; Volkert, W.
A.; Ketring, A. R. Phosphorus, Sulfur Silicon 1993, 77, 267. (c) Wang,
M. F.; Volkert, E. W.; Singh, P. R.; Katti, K. K.; Lusiak, P.; Katti, K.
V.; Barnes, C. L. Inorg. Chem. 1994, 33, 1184. (d) Singh, P. R. New
(14) (a) Keana, J . F. W.; Cai, S. X. J . Org. Chem. 1990, 55, 3640. (b)
Keana, J . F. W.; Cai, S. X. J . Fluorine Chem. 1989, 43, 151. (c) Cai, S.
X.; Glenn, D. J .; Keana, J . F. W. J . Org. Chem. 1992, 57, 1299. (c)
Keana, J . F. W.; Cai, S. X. J . Fluorine Chem. 1989, 43, 151. (d) Cai, S.
X.; Keana, J . F. W. Tetrahedron Lett. 1989, 30, 5409. (e) Cai, S. X.;
Glenn, D. J .; Gee, K. R.; Yan, M.; Cotter, R. E.; Reddy, N. L.; Weber,
E.; Keana, J . F. W. Bioconjugate Chem. 1993, 4, 545.
(
15) Pandurangi, R. S.; Katti, K. V.; Barnes, C. L.; Volkert, W. A.;
Kuntz, R. R. Chem. Commun. 1994, 1841.
16) (a) Pandurangi, R. S.; Kuntz, R. R.; Volkert, W. A. J . Nucl. Med.
B), Int. J . App. Rad. Isotopes 1995, 46, 233. (b) Pandurangi, R. S.;
Karra, S. R.; Kuntz, R. R.; Volkert, W. A. Bioconjugate Chem. 1995,
(5), 630.
17) Pandurangi, R. S.; Karra, S. R.; Kuntz, R. R.; Volkert, W. A.
Photochem. Photobiol. 1996, 64(1), 100.
18) (a) Pinney, K. G.; Carlson, K. E.; Katzenellenbogen, B. S.;
(
(
6
(
9
9m
186
109
Tc, Re and Pd Complexes of SN and PN Ligand systems, Ph.D.
Dissertation, University of MissourisColumbia, 1992. (e) Wang, M.
(
9
9m
109
Katzenellenbogen, J . A. J . Biochem. 1991, 30, 2421. (b) Pinney, K. G.;
Katzenellenbogen, J . A. J . Org. Chem. 1991, 56, 3125. (c) Bergmann,
K. E.; Carlson, K. E.; Katzenellenbogen, J . A. Bioconjugate Chem. 1994,
F. New
Tc and Pd/ Pd complexes with P-N Ligands, M. S.
Dissertation, University of MissourisColumbia, 1993. (f) Katti, K. V.;
Ge, Y. W.; Singh, P. R.; Date, S. V.; Barnes, C. L. Organometallics
1994, 13, 541.
5
, 141.

2
800 J . Org. Chem., Vol. 62, No. 9, 1997
Pandurangi et al.
Sch em e 1
Sch em e 2
Sch em e 3
base adduct 2 in good yields. The imine functionality in
was reduced to 3 in high yields with NaBH CN. The
2
3
transformation of 2 to 3 is intended to increase the in
vitro and in vivo stability of the hydrazido framework.
In addition, the introduction of an NH functionality in 3
is expected to offer suitable electron donor characteristics
for interactions with Pd(II) and other metallic radioiso-
topes. The monofunctionalized organophosphorus hy-
drazides 2 and 3 undergo coupling with 4 at the other
terminal hydrazido center to produce the azidoperfluo-
roaryl amides 5 and 6, respectively, as described in
Scheme 2. The perfluoroaryl azido-functionalized phos-
phorus hydrazides 5 and 6 represent the first examples
of main group ligating systems capable of undergoing
photochemically induced CH (or NH) insertion reactions.
tion with elemental analysis data further confirm the
chemical constitutions of these compounds.
Meta la tion Rea ction of 6 w ith P d Cl
reaction of 6 with PdCl (PhCN) was carried out in CH
Cl at 25 °C to produce 7 (Scheme 3). Complex 7 is a
2 2
(P h CN) . The
1
New compounds 3, 5, and 6 were characterized by H,
2
2
2
-
3
1
19
P, and F NMR spectroscopy and C, H, and N elemen-
2
1
tal analysis. H NMR spectra of the monosubstituted
Schiff base adducts 2 and its reduced form 3 consisted of
two doublets for the NMe protons (J P-H ) 9-12 Hz),
indicating the presence of two dissimilar hydrazine
backbones. The IR spectrum of 6 shows bands at 1689
cm and 3267 cm assigned to ν CdO stretching and ν
NsH stretching modes, respectively, confirming the
formation of the amide bond. The 31P NMR spectroscopic
data indicate modest upfield shifts (3-4 ppm) for 2, 3,
dark-brown, air-stable, microcrystalline solid. The in-
teraction of Pd(II) with 6 can occur in two possible
ways: (i) one that involves the coordination of the two
hydrazines (structure A), and (ii) via the coordination of
the reduced amino (CH NH ) arm of the hydrazine in 6
2 2
in conjunction with the P(S) functionality (structure B)
-
1
-1
as shown below.
The 1H NMR spectrum of 7 consisted of features
diagnostic of structure B. For example, the CH
tionality of the reduced nitroaromatic arm of 6 resonated
as two inequivalent nuclei (H and H ) centered at δ 3.76
2
func-
5
, 6 as compared to the parent 1. It is also of note that
the magnitude of J P-H values (due to H CNP(S)) are 2-3
3
A
B
Hz lower as compared to 1. This observation may be
rationalized in terms of an electronic effect which,
and 4.04, respectively. These two resonances underwent
a downfield shift to δ 4.12 and 4.85 in 7 suggesting that
NH nitrogen attached to CH2 functionality is involved
in the complexation to Pd(II). The ν CdO stretching
presumably, decreases the s-character across the PNCH
3
skeleton upon introduction of aryl or the perfluoroaryl
substituents at the terminal nitrogens as outlined in
Schemes 1 and 2. The IR spectroscopic data in conjunc-
-
1
band at 1685 cm observed for 7 is nearly identical to
that found for 6, indicating the lack of interaction of

Chemistry of Bifunctional Photoprobes
J . Org. Chem., Vol. 62, No. 9, 1997 2801
Sch em e 4
carbonyl moiety with the Pd(II) center. The ν PdS in 6
-
1
was observed at 622 cm . This frequency was shifted
-
1
to 598 cm
in 7, indicating the coordination of the
phosphorus chalcogenide with the Pd(II) center. Similar
shifts in PdS stretching frequencies for a number of
metal complexes of phosphorus hydrazides where N, S
coordination with Pd(II) as depicted in structure B have
been noted.21 In fact, the N, S coordination as depicted
for 7 has been demonstrated for a number of Pd(II)
complexes and also confirmed by X-ray crystallogra-
Ta ble 1. Absor p tion Ma xim a a n d Reten tion Tim es for
P er flu or oa r yl Azid o P h osp h or u s Hyd r a zid es in
Cycloh exa n e
phy.2
1,22
-1
The intense absorption around 2120-2130 cm
in compounds 5, 6, and 7 is assigned to N
3
and indicates
a
that the azido group is not perturbed or involved in the
functionalization process. C, H, and N analysis of 7
reveals only one phosphorus hydrazide ligand per metal
center.
In ser tion Rea ction s of P er flu or oa r yl Azid o P h os-
p h or u s Hyd r a zid es. The insertion efficiency of ni-
trenes generated by the photolysis of tetrafluoroaryl
azides is known to be very sensitive to the substitution
in the position para to the azide.23 For example, electron-
retention time (tR/min)
λmax
before
after photolysis
compound
(nm) photolysis (C-H insertion adduct)
4
5
6
7
-N3-C6H4-COOCH3 260
0.8
2.1
2.2
1.2
2.1
4.7
4.9
1.9
265
266
267
a
CH CN:H O in the ratio 2:1 at a flow rate of 1 mL/min.
3
2
withdrawing groups (e.g., NO
are known to enhance nitrene insertion into cyclohexane
as compared to electron-donating groups (e.g., NH , OH,
2
, CN, COCl, COOMe, etc.)
yielded 9 which upon refluxing with SOCl
acid chloride-cyclohexane adduct 10. Treatment of 10
with 2 and 3 in the presence of Et N resulted in the
2
yielded the
2
3
1
etc.). Hence, azides used for labeling experiments should
be evaluated for singlet nitrene insertion reactions with
secondary amines to form substituted hydrazines (NH
insertion) or cyclohexane (CH insertion) to give cyclohexyl
desired cyclohexane adducts 11 and 12. H NMR of 12
shows characteristic peaks at δ 4.04 and 3.78 which have
been assigned to NH and CH protons of the nitrene
adduct, respectively, from previous studies on similiar
systems.15 The assignment was based on deuterium
exchange where the δ 4.04 peak disappeared on addition
derivatives. Photolysis of 5 and 6 in a mixture of CH
Cl and cyclohexane (1:1, v/v) produces an insertion
2
-
2
adduct through the activation of a C-H bond as moni-
tored by HPLC (Scheme 4). The insertion products elute
2
of D O. Assignment of the methine carbon was further
1
3
confirmed by C NMR using the DEPT sequence which
distinguished the methine carbon from the methylene
carbons of cyclohexane. The synthesized 12 shows the
same NMR spectral features and elutes at the same time
at t
min for the parent azides 5 and 6, respectively (Table
). In order to confirm the structure of the adduct and
R
) 4.7 and 4.9 min as compared to t ) 2.1 and 2.2
R
1
estimate the efficiency of insertion of the azides 5 and 6
onto the cyclohexane, a route was developed to synthesize
the cyclohexane adduct of the perfluoroaryl azido phos-
phorus hydrazide as described in Scheme 5. HPLC
retention time of the synthetic product was used to
identify the inserted adduct among the photolyzed prod-
ucts and, subsequently, as a standard for quantitative
estimation. Methyl 4-azido-2,3,5,6-tetrafluorobenzoate
which is known for insertion onto cyclohexane was
photolyzed in cyclohexane for 5 h and chromatographi-
R
(t ) 4.9) as the major photolyzed product, identifying
the photolytically produced cyclohexane adduct.
A significant outcome of this investigation is the
determination of CH insertion efficiency of the nitrenes
produced by photolysis of the perfluoroaryl azido phos-
phorous hydrazides in cyclohexane. Photolysis of dilute
-
4
solutions (2 × 10 M) of 5 and 6 in cyclohexane and CH
2
-
Cl mixtures gave >90% destruction of the parent azide
2
in 2 min photolysis with insertion product yields of 78 (
3% for 5 and 81 ( 5% for 6 as determined by HPLC
analysis. The short exposure times, consistent with a
cally separated to yield the CH insertion adduct 8 as
reported earlier.1
2a
Hydrolysis of 8 using 30% NaOH
15,16a
high quantum efficiency for photoprobe destruction,
are of significance for potential photolabeling of proteins.
Long exposure could directly damage the proteins and
increases the possibility of secondary photolysis leading
to decomposition of photoproducts or sensitized damage
to the protein. Hence, these two azides provide the first
(
22) Pandurangi, R. S.; Kuntz, R. R.; Volkert, W. A.; Barnes, C. L.;
Katti, K. V. J . Chem. Soc., Dalton Trans. 1995, 565.
23) Pandurangi, R. S.; Katti, K. V.; Volkert, W. A.; Kuntz, R. R.
(
International Conference on Photochemistry, Proceedings; Vancouver,
Canada, 1993; p 367.

2
802 J . Org. Chem., Vol. 62, No. 9, 1997
Pandurangi et al.
Sch em e 5
Sch em e 6
azide moiety are useful for identifying different photo-
products. Typical photochemical reactions of perfluoro-
aryl azides involve generation of singlet products (bond
insertion) and triplet products (noninsertion products
2
4
including amines and azo compounds).
The sym-
metrical AA′XX′ pattern for 19F NMR of perfluoroaryl
azides could be useful in the identification of ring-
expanded products which often plague the photochem-
istry of aryl azides.1 Compound 6 in cyclohexane shows
two multiplets δ -139.4 and -151.9 corresponding to F-2,
F-6 and F-3, F-5, respectively (Figure 1a), before pho-
tolysis. Partial photolysis of the mixture in the NMR
tube (Figure 1b) results in the development of two
multiplets at δ -141.1 and -161.9. When triplet prod-
ucts were preferentially enhanced by photolyzing the
mixture in the presence of a triplet sensitizer (acetophe-
none), two additional multiplets at δ -141.8 and -163.3
appear (Figure 1c) which can be assigned to the amine
produced through hydrogen abstraction by triplet nitrene.
An independently synthesized 12 shows 19F NMR ab-
sorptions at δ 140.7 and 162.1 (Figure 1d) coinciding with
the cyclohexane adduct produced photolytically. Thus,
0a
1
9
F NMR can be an effective tool in differentiating
insertion and noninsertion products. The relative inte-
gration of fluorine signals with respect to total fluorine
integration reflects the relative efficiency of insertion vs
noninsertion. The two multiplets of low intensity at δ
example of a system in which a large molecular frame-
work involving a complexing moiety shows high efficiency
insertion reactions provided it is suitably tuned. This
superior yield has a high impact on the utility of such
chromophores for efficient photoattachment to different
macromolecular probes. Similarly, the photolysis of 7 in
1
50.5 and 138.3 (Figure 1c) can be assigned to the azo
compound formed by the recombination of triplet ni-
15
trenes. Photolysis of 6 in methanol gives exclusively
triplet products (azo and amine derivatives, not shown
19
in Figure 1), the F NMR of which is similar to that of
2 2
a cyclohexane and CH Cl mixture (1:1, v/v) resulted in
triplet sensitized products. It is well established that
a 45 ( 3% CH insertion adduct yield (Scheme 6) as
determined by 19F NMR and HPLC analysis based on
an authentic standard obtained by independent synthe-
sis. Although the introduction of Pd metal into the
hydrazine framework reduced the CH insertion yield, the
complex still may conjugate to proteins with high ef-
ficiency by virtue of the close proximity of the ligand and
the protein induced by hydrophobic interactions between
the probe and hydrophobic pockets prevailing in
methanol catalyzes the intersystem crossing leading to
15
13c
triplet products.
The differentiation between insertion
and noninsertion products, especially the amines, can
19
19
also be achieved by F- F homodecoupling procedures.
Decoupling of F-2 and F-6 in the photolyzed mixture
containing cyclohexane adduct and amine by irradiating
at the center frequency, δ -141.1, changes the coupling
pattern of the peak at δ -161.9 to a simple doublet
1
corresponding to H coupling (J ) 3.5 Hz) in the C-H
1
6a
proteins.
insertion adduct. For the amine, irradiating at the center
In Situ 19F NMR Sp ectr oscop y of P h otoin ser tion
frequency δ -141.8 resulted in the collapse of the
1
9
vs Non in ser tion P r od u cts. The advantage of F NMR
spectroscopy for the study of fluorinated biomolecules
originates from the high sensitivity, wide chemical shift
1
multiplet at δ -163.7 to a triplet corresponding to H
coupling (J ) 4.5 Hz). This decoupling procedure of
1
9
range and high abundance of the F nucleus. In
particular, the chemical shifts of fluorines flanking the
(24) Hibert, F. K.; Kapfer, I.; Goeldner, M. Angew. Chem., Int. Ed.
Engl. 1995, 34, 1296.

Chemistry of Bifunctional Photoprobes
J . Org. Chem., Vol. 62, No. 9, 1997 2803
F igu r e 2. Size exclusion chromatogram (SEC-HPLC) of 1
F igu r e 1. 19F NMR spectrum of 6 in (a) cyclohexane (before
photolysis); (b) photolysis in NMR tube for <5 min; (c)
photolysis in the presence of triplet sensitizer (acetophenone);
-5
with HSA (both 4 × 10 M) in pH 6.8 buffer: (a) before
photolysis, (b) 15 s exposure, (c) 30 s exposure, (d) after
treatment with SDS. The eluting solvent was 0.02 M monoba-
sic sodium phosphate buffer containing 0.05 M sodium sulfate
1
9
(d) F NMR of independently synthesized cyclohexane adduct.
(pH 6.8) at a flow rate of 1 mL/min. Relative detector response
refers to division of pen deflection on a strip chart recorder.
identifying insertion and noninsertion products is quite
useful when fluorinated probes are attached to proteins
where the photoproducts are difficult to isolate and
characterize.
The distribution of products from photolysis of per-
fluoroaryl azides depends on the reactivity of the singlet
nitrene and the environment in which it is generated.
In a constrained medium (e.g., liquid crystalline or matrix
isolated materials) even triplet nitrene can produce
formal insertion products through abstraction of hydro-
gen from the substrate followed by radical combination.
The noninsertion products are readily separated from
native or photoconjugated proteins under SEC-HPLC
conditions by virtue of the great disparity in molecular
mass. For example, the triplet nitrene-derived amine
derivative from photolysis of 6 in buffer elutes at 11.5
min under the HPLC conditions employed (not shown in
Figure 2). The absence of this product, even after SDS
denaturation of HSA, is consistent with a high efficiency
Con ju ga tion of 6 w ith HSA. Equimolar solutions
of HSA and 6 in aqueous buffer at pH 6.8 were incubated
for 1 h before injecting into a SEC-HPLC column. Before
photolysis, HSA elutes at 8.2 min and the photoprobe 6
at 10.8 min (Figure 2a). This separation demonstrates
that the probe is not bound to HSA prior to photolysis.
Photolysis (Figure 2b,c) results in disappearance of the
probe accompanied by an increase in the absorbance in
the protein peak. Denaturing of the photolyzed protein-
photoprobe mixture with sodium dodecyl sulfate (SDS)
has no effect other than dilution on the chromatogram
features (Figure 2d). Although these absorbance changes
do not constitute a proof of covalent attachment, they are
consistent with the similar behavior seen in the photo-
1
6b
for covalent attachment of the probe to the protein.
Since covalent binding of nitrene-based photoprobes to
labeling of HSA, IgG,1 and B72.3 where covalent
6b
17
proteins usually exceeds the C-H bond insertion ef-
bonding was confirmed with 14C-labeling.
ficiency on cyclohexane,
16b,17
a high insertion efficiency

2
804 J . Org. Chem., Vol. 62, No. 9, 1997
Pandurangi et al.
of 109Pd activity with HSA was investigated by denatur-
ing the protein with sodium dodecyl sulfate (SDS) which
is expected to open the tertiary structure of the protein
and release noncovalently bound molecules. Under these
conditions (Figure 3B) there is no significant loss of
activity in the HSA region, and we conclude the label is
1
6
covalently attached to the protein. Previously, it was
found that the amine product from the triplet nitrene
precursor is the only major side photoproduct in the
absence of protein. Photolysis of 7 in the absence of HSA
shows that this noninsertion product elutes at a similar
retention time (t
R
) as that of the unphotolyzed probe
(Figure 3B). Although the resolution is inadequate to
resolve 7 and the amine photoproduct, it does show that
both are separable from the labeled HSA. Peak broad-
ness, due to the eluent collection procedure, also pre-
cluded complete separation of the HSA and probe frac-
tions of the chromatogram. However, if fractions 7-9
in Figure 3 are assumed to be exclusively protein and
its labeled products, it can be estimated that 25% of the
1
09
Pd activity is bound to the protein after 1 h incubation
before photolysis. Following a 3 min photolysis, 74% of
the activity remains with the protein, and this value is
still 71% after denaturation with SDS. Whether or not
this preassociated activity becomes covalently bound to
HSA during photolysis cannot be ascertained. Neverthe-
less, the 49% increase in labeling following photolysis sets
a lower limit for the photoconjugation efficiency, with a
possible efficiency as high as 74% if the preassociated
F igu r e 3. SEC-HPLC chromatograms of the 109Pd complex
7
: (A) with HSA before photolysis (b), after 1 min photolysis
(9), after 3 min photolysis (2); (B) treatment of the 3 min
photolysis mixture with SDS (b), and after photolysis of 7 in
buffer without HSA (9).
109
Pd complex is also covalently bound by photolysis.
19
F NMR Sp ectr oscop y of th e P h otocon ju ga tion
of 6 w ith HSA. Application of 19F NMR to distinguish
of the singlet nitrene derived from 6 in HSA is consistent
with the observed 82% efficiency for insertion into
cyclohexane.
between insertion and noninsertion reactions in proteins
19
was attempted with 6 and HSA. The F NMR spectrum
of 6 (dissolved in a minimum amount of ethanol) and
HSA in phosphate buffer shows two multiplets at δ
1
09
Com p lexa tion of P d w ith 6. The complexation of
2
-
109
the activated PdCl
carried out in ethanolic solution. The complex, isolated
by extraction into CHCl and washing with saline, was
4
sample containing Pd with 6 was
-
144.6 (F-2, F-6) and δ -151.9 (F-3, F-5, Figure 4a). A
partial photolysis of this mixture produced two new
3
multiplets at δ -146.4 and -161.2 (Figure 4b). The δ
characterized by paper chromatography as a neutral
complex with a radiochemical purity of >98%.
P h ot ocon ju ga t ion of t h e 109P d -La b eled 7 w it h
-
161.2 multiplet matches the δ -162.1 feature observed
for F-3 and F-5 for the independently synthesized cyclo-
hexane adduct (Figure 1d). This observation may be
rationalized as an insertion of the photoprobe into
HSA. The total concentration of the Pd complex stock
1
09
25
solution (“cold” Pd and
Pd) is estimated to be 4.95 ×
, assuming the 75% yield of complexation as found
for unactivated Pd. Human serum albumin (HSA) (5 mL,
hydrophobic regions of HSA. To confirm this hypoth-
-
3
1
0
esis, the experiment was repeated in the absence of HSA
where the only photoproduct is expected to be the triplet-
derived amine. The spectrum of the prephotolysis mix-
ture appears in Figure 4c. After photolysis (Figure 4d),
a new multiplet appears at δ -163.4 which corresponds
to the chemical shift of the triplet nitrene-derived amine.
Although the difference in chemical shifts of F-3 and F-5
between the insertion and amine products is quite small
(δ 1.3) it is adequate for differentiation by in situ NMR
studies. However, the low signal to noise ratio and peak
broadness precludes a quantitative evaluation of inser-
tion efficiency by this method.
-
5
109
2
× 10 M) was mixed with of the
µL, ∼5 × 10 M in total Pd complex) and incubated for
h. This solution with an approximate ratio of 7/HSA
Pd complex (100
-
3
1
of 5/1 was divided with part used for a control and part
for the photoconjugation experiment. Samples (20 µL)
of the mixture were injected onto the SEC-HPLC. Prior
to photolysis, the photoprobe elutes between 10 and 11
min (Figure 3A) indicating the 1 Pd complex is separable
from HSA which elutes at 7-9 min on this column. The
small amount of activity appearing in the HSA region of
the chromatogram is indicative of a minor prephotolysis
binding of the 1 Pd complex to HSA, but the major
fraction remains unattached to the protein before pho-
tolysis. The nature of the preassociation has not been
elucidated. However, 31P NMR of 7 in the presence of
HSA indicates a single compound analogous to that
09
The chemical shifts of F-2 and F-6 are quite sensitive
to the presence of HSA. In cyclohexane and in buffer
without HSA, these resonances occur at δ -139.4 and
-140.7, respectively, while in buffer with HSA present,
the resonance occurs at δ -144.6. Although the origin
of this shift is unclear, fluorine chemical shifts are known
09
2
6
observed for 7 in CDCl
3
, suggesting that the Pd-chelating
to be very sensitive to environment.
For example,
Fenney et al.27 have reported large changes of chemical
environment is not altered by the association. Sequential
photolysis (1-3 min) of the mixture shifts a major
fraction of the radioactivity to the HSA region of the
chromatogram (Figure 3A). The nature of the association
(
25) Spector, A. A.; J ohn, K.; Fletcher, J . J . Lipid Res. 1969, 10, 56.
(26) Lau, E. Y; Gerig, J . T. J . Am. Chem. Soc. 1996, 118, 1194.

Chemistry of Bifunctional Photoprobes
J . Org. Chem., Vol. 62, No. 9, 1997 2805
insertion yield (78% and 82% for 5 and 6, respectively)
on cyclohexane. The high quantum yields for photoprobe
binding is of considerable significance in photolabeling
of proteins since long exposure may not be warranted for
1
9
labeling biomolecules. In situ F NMR provides a
convenient method for differentiation between insertion
and noninsertion products in proteins and model sol-
vents.
Despite some prephotolysis binding between HSA and
, the 1 Pd-labeled 7 provides a convenient radiochemical
09
7
probe to investigate the photoconjugation of proteins. The
high efficiency for covalent binding to HSA and the small
yields of the polar noninsertion photoproduct coupled
with the facile separation of the covalently bound radio-
labeled heterobifunctional chelating agent from the non-
covalently bound photoproducts greatly enhances the
potential applications of photolabeling in nuclear medi-
cine. Although triplet nitrene products dominate when
azides are photolyzed in hydroxylic media, these appear
not to be significant in the presence of HSA. Our
hypothesis is that the hydrophobic photoprobe is pro-
tected from solvent exposure by association with the
hydrophobic regions of the protein. The extent to which
this protection will continue with less hydrophobic pro-
teins remains to be investigated.
Exp er im en ta l Section
All synthetic procedures were conducted in a dry nitrogen
atmosphere using standard Schlenk tube techniques and
prepurified solvents. Reactions involving the synthesis of
azides were carried out in subdued light by wrapping the flasks
with aluminum foil. Phosphorus hydrazide 1 was prepared
as reported earlier.2 NMR spectra were recorded in CDCl
1b
3
on a 300 MHz spectrometer, and chemical shifts are reported
31
in ppm downfield from SiMe
chemical shifts are reported with respect to 85% H
external standard, and positive shifts lie downfield from the
4
for 1H NMR. The P NMR
3
PO as an
4
1
9
standard. F NMR chemical shifts are reported with respect
to CFCl as an external standard. Infrared spectra were
3
recorded in KBr pellets. Elemental analysis for the new
compounds were done by Oneida Research Services, Inc., New
York. UV/visible spectra were recorded either in cyclohexane
or in acetonitrile.
Protein assays were performed by injecting 20 µL samples
onto a size exclusion column (Biorad Biosil SEC 250). The
eluent was a phosphate (PBS) buffer (0.02 M sodium phos-
phate, monobasic containing 0.05 M sodium sulfate, pH 6.8)
pumped at the rate of 1 mL/min. Experiments performed with
HSA and the bifunctional ligand before 1 Pd labeling were
analyzed by 254 nm UV detection of the eluent. For mixtures
of HSA and the radiolabeled complex, the eluent was collected
in 1 mL fractions and subsequently counted with a NaI
counter.
Photolysis was carried out with a 200 W high pressure Hg
lamp. The focused beam passed through a water filter and a
glass cut-off filter to limit excitation to wavelengths greater
than 320 nm. Compounds 5-7 have similar absorption
spectra with λmax near 265 nm (Table 1), diminishing to <10%
of the λmax value at 320 nm. Thus, photolysis was carried out
in the spectral region where proteins have minimal absor-
bance. A standard silica cuvet was utilized for labeling
experiments which typically required 2 min irradiations for
F igu r e 4. 19F NMR of 6 with HSA (a) before photolysis
control, (b) after partial photolysis, (c) 6 in buffer without HSA
before photolysis, (d) after 3 min photolysis of c. All 19F NMR
3
chemical shifts are referenced to CFCl .
09
shifts in 19F-labeled proteins which may be due to
conformational as well as local field effects. The prefer-
ential small shift of F-2 and F-6 (δ 3.9) may be related
to the prephotolysis orientation of the probe in the
hydrophobic pocket of the protein. The much larger
shifts of F-3 and F-5 after photolysis, however, are a clear
indication of covalent modification of the protein with the
photoprobe.
Con clu sion
Incorporation of perfluoroaryl azides into phosphorus
hydrazide ligating systems represent a new approach for
the formulation of versatile bifunctional chelating agents
carrying both photoactivable and chelating moieties. The
amide-coupled phosphorus hydrazide derivatives appear
to possess global reactivity by showing a high C-H
>
90% completion. In situ experiments were performed by
placing the NMR tube such that it intersected a portion of the
light beam. The in situ studies typically required 10 min for
>
90% destruction of the photoprobe due to the small fraction
of the beam area intercepted by the NMR tube. The photolytic
decomposition of the photolabels was monitored by HPLC
using a 30 mm × 4.7 mm C-18 reverse phase column. t
R
values
are reported in minutes using eluent CH CN-H O in ratio
3
2
(27) Feeney, J .; McCormick, J . E.; Bauer, C. J ., Birdsall, B. Moody,
C. M.; Starkman, B. A.; Young, D. W.; Francis, P.; Halvin, R. H.;
Arnold, W. D.; Oldfield, E. J . Am. Chem. Soc. 1996, 118, 8700.
2:1 at a flow rate of 1 mL/min. Solutions were purged with
prepurified nitrogen for several minutes before photolysis.

2
3
806 J . Org. Chem., Vol. 62, No. 9, 1997
Pandurangi et al.
Syn t h esis of (C
6
H
5
)P (S)(NMeNH
2
)(NMeNHCH
2
-C
6
H
4
-
were separated by column chromatography using hexane, ethyl
-NO ) (3). Monofunctionalized phosphorous hydrazide 2 (2.8
2
acetate, and methanol in the ratio of 20:1:1. The amine 8 (0.09
2
2
1
g, 7.7 mmol) prepared earlier and a trace of bromocresol
green were dissolved in CH OH (50 mL), and NaBH CN (2.9
g, 46.2 mmol) was added. The solution immediately turned
g, 57%) crystallized upon cooling. Mp: 85 °C; H NMR: δ 1.22
3
3
(m 3H) 1.31 (m 2H); 1.58 (m, 1H), 1.76 (m, 2H), 2.03 (m, 2H),
3.61 (m, 1H), 3.91 (s, 3H), 4.02 (m, 1H); 19F NMR: δ -144.2
(m, 2F), -161.2 (m, 2F); IR: 3368, 3007, 2930, 1714, 1638,
deep blue. The pH of the solution was adjusted to ∼4.0 by
-
1
adding 2 N HCl-CH
3
OH dropwise with stirring until the
1541, 1311, 1082, 881, 742 cm . Anal. Calcd for C14
NO C, 55.08; H, 4.95; N, 4.59. Found: C, 55.21; H, 4.72; N,
4.49.
15 4
H F -
yellow color was restored. After 6-8 h, NaBH
mmol) in methanol was added and the solution stirred
overnight. The excess NaBH CN was destroyed with 5 N HCl.
4
The product was extracted into CH Cl and dried with MgSO ,
3
CN (1.93 g, 30.8
2
3
In order to prepare adequate quantities of 8 for syntheses
of other cyclohexylamino derivatives an independent synthetic
procedure was also employed.1
2
2
6b
and the solvent was evaporated in vacuo to yield a pale yellow
crystalline solid (2.39 g, 85%). Mp: 104-106 °C; IR: 3309,
Syn t h esis of 4-(Cycloh exyla m in o)-2,3,5,6-t et r a flu o-
r oben zoic Acid (9). The cyclohexane adduct 8 (5 g, 16.3
1
3
225, 3056, 2943, 1520, 1428, 1337, 1104, 999, 717 cm-1; H
NMR: δ 2.76 (d, J ) 12.37 Hz, 3H), 3.01 (d, J ) 10.34 Hz,
mmol) was dissolved in CH OH (100 mL), and the adduct was
3
3
4
4
4
H), 3.40 (br s, 2H), 3.89 (d, J ) 12.51 Hz, 1H), 4.04 (br, 1H),
hydrolyzed using 30% aqueous NaOH (25 mL) by stirring
overnight at 25 °C. The solution was acidified by 2 N HCl at
.07 (d, J ) 12.21 Hz, 1H), 7.42-7.55 (m, 5H), 8.02-8.10 (m,
3
1
H); P NMR: δ 85.2. Anal. Calcd for C15
H
20
N
5
O
2
PS: C,
0 °C to pH < 1 and extracted into CHCl layer (3 × 30 mL)
3
9.31; H, 5.52; N, 19.17. Found: C, 49.21; H, 5.34; N, 19.27.
and dried using MgSO and evaporated to a white powder 9
4
Syn th esis of 4-Azid o-2,3,5,6-tetr a flu or oben zoyl Ch lo-
(85%). Mp: 163-165 °C; IR: 3389, 2937, 1687, 1638, 1527,
-
1 1
r id e (4). A mixture of 4-azidotetrafluorobenzoic acid (5 g,
1.18 mmol) and excess SO Cl (10 mL) in dry hexane (250
1305, 1249, 985, 763, 714 cm ; H NMR: δ 1.20 (m, 3H), 1.32
(m, 2H), 1.57 (m, 1H), 1.75 (m, 2H), 2.04 (m, 2H), 3.60 (m,
1H), 4.04 (m, 1H), 11.5 (s, 1H); 19F NMR: δ -139.5 (m, 2F),
2
2
2
mL) was refluxed until the acid was dissolved completely,
indicating the completion of the reaction. The solvent was
removed, and the residue was distilled (100 °C/0.5 mmHg)
using a K u¨ gelrohr apparatus to give a yellow liquid (79%). The
physical constants of the compound coincided well with the
-165.1 (m, 2F). Anal. Calcd for C13
13 4 2
H F NO C, 53.59; H,
4.50; N, 4.81. Found: C, 53.12; H, 4.43; N, 4.76.
Syn t h esis of 4-(Cycloh exyla m in o)-2,3,5,6-t et r a flu o-
r oben zoyl Ch lor id e (10). A mixture of 4-(cyclohexylamino)-
2,3,5,6-tetrafluorobenzoic acid (9) (400 mg, 1.37 mmol) and
1
4a
earlier report.
Syn t h esis of (C
CHC -3-NO ) (5). A solution of 4 (910 mg, 3.59 mmol) in
CH Cl was slowly added dropwise to the solution of 2 (1 g,
.75 mmol) in CH Cl at 0 °C in the presence of excess Et
560 mg, 5.53 mmol) and stirred at 25 °C for 2 h. The mixture
6
H
5
)P (S)(NMeNH COC
6
F
4
N
3
)(NMeNd
2 2
SO Cl (1.5 mL) in dry hexane (35 mL) was refluxed for 1 h.
6
H
4
2
The solvent was removed, and the crude acid chloride 10 was
used as such for preparation of the cyclohexane adduct 11.
IR: 3400, 2921, 2851, 1752, 1639, 1541, 1301, 984, 738, 710
2
2
2
(
2
2
3
N
-
1
1
cm ; H NMR: δ 1.19 (m, 3H), 1.33 (m, 2H), 1.55 (m, 1H),
19
was washed with water (25 mL) and brine (25 mL) after which
it was dried using sodium sulfate. The solvent was evaporated
in vacuo, and the crude product was purified on a silica gel
column by eluting with 30% ethyl acetate-hexane buffered
1.74 (m, 2H), 2.04 (m, 2H), 3.61 (m, 1H), 4.11 (m, 1H);
F
NMR: δ -143.6 (m, 2F), -166.4 (m, 2F).
Syn th esis of Cycloh exa n e Ad d u ct 11. To the entire
sample of 10 dissolved in CH Cl (5 mL) was added a mixture
2
2
with 1% Et
3
N to give the amide 5 (1.28 g, 80%). Mp: 60-63
of 2 (300 mg, 0.82 mmol) and Et N (170 mg, 1.68 mmol) in
3
°
7
C; IR: 3069, 2914, 2126, 1700, 1647, 1533, 1486, 1349, 985,
dry CH Cl2 (10 mL) at 0 °C under a N2 atmosphere. The
2
-
1 1
87, 734 cm ; H NMR: δ 3.09 (d, J ) 10.36 Hz, 3H), 3.46 (d,
reaction mixture was allowed to come to room temperature,
J ) 8.87 Hz, 3H), 7.4-8.2 (m, 9H), 7.80(s,1H), 8.28 (m, 1H);
stirred for 2 h, diluted with CH Cl2 (100 mL), washed with
2
3
1
P NMR: δ 81.8; 19F NMR: δ -140.8 (m, 2F), -150.4 (m, 2F).
Anal. Calcd for C22 PS: C, 45.52; H, 2.95; N, 19.30.
Found: C, 45.36; H, 3.21; N, 18.97.
Syn t h esis of (C )P (S)(NMeNH COC
HCH -3-NO ) (6). The procedure was similar to that
used for the preparation of 5. A CH Cl solution of 4 (2.08 g,
.1 mmol) and 3 (1.92 g, 5.25 mmol) reacted to give 6 (2.51 g,
water (20 mL) and brine (20 mL), and dried over anhydrous
sodium sulfate. Removal of the solvent under reduced pres-
sure yielded an oil which was purified on a silica gel column
by eluting with 20% ethyl acetate-hexane buffered with 1%
17 8 4 3
H N F O
6
H
5
6 4 3
F N )(NMeN-
2
C
6
H
4
2
Et
3
N to give 11 (392 mg) in 74% yield. Mp: 65-66 °C; IR:
2
2
3337, 3062, 2929, 2815, 1689, 1647, 1499, 1344, 1146, 956, 731,
-
1 1
8
8
2
5
9
674 cm ; H NMR: δ 1.07-1.42 (m, 5H), 1.60-1.76 (m, 3H),
1.95-2.02 (m, 2H), 3.09 (d, J ) 10.19 Hz, 3H), 3.43 (d, J )
2%) as a pale yellow solid. Mp: 95-96 °C; IR: 3267, 3055,
858, 2126, 1689, 1639, 1520, 1337, 1252, 984, 731, 682, 622,
8.89 Hz, 3H), 3.56 (m, 1H), 3.93 (m, 1H), 7.44-7.55 (m, 4H),
-
1
1
31
19 cm ; H NMR: δ 2.83 (d, J ) 11.12 Hz, 3H), 3.14 (d, J )
7.67 (s, 1H), 7.81-8.1 (m, 4H), 7.84(s, 1H), 8.26 (m, 1H);
P
1
9
.96 Hz, 3H), 3.76 (d, J ) 11.87 Hz, 1H), 4.01(br, 1H), 4.04 (d,
NMR: δ 80.7; F NMR: δ -143.8 (m, 2F), -161.1 (m, 2F);
1
3
J ) 12.33 Hz, 1H), 7.40-7.48 (m, 5H), 7.65 (s, 1H), 7.95 (s,
C NMR: 157.7, 148.4, 146.9(m), 143.0(m), 138.1(m), 137.0,
19
31
13
31
1
(
4
H), 8.07-8.15 (m, 3H); 31P NMR: δ 83.8; F NMR: -140.6
135.4(d) [J ( P- C), 11.9 Hz], 134.4(m), 132,2, 132.0 [J ( P-
13
m, 2F), -150.2 (m, 2F). Anal. Calcd for C22
H
19
N
8
F
4
O
3
PS: C,
C), 3.7 Hz], 131.6, 129.6, 128.2, 128.0, 122.9, 121.7, 53.4 [CH,
1
3
19
31
13
31
5.37; H, 3.29; N, 19.24. Found: C, 45.69; H, 3.29; N, 19.49.
J ( C- F), 4.8 Hz], 37.5 [J ( P- C), 7.4 Hz], 34.2, 31.8 [J ( P-
13
Syn th esis of P a lla d iu m Com p lex 7. A solution of
C), 10.2 Hz], 25.3, 24.5. Anal. Calcd for C28
C, 52.83; H, 4.59; N, 13.20. Found: C, 52.62, H, 4.54, N, 13.29.
HPLC t 4.7 min.
Syn th esis of Cycloh exa n e Ad d u ct 12. The procedure for
11 was followed using 3 (75 mg, 0.27 mmol), Et N (60 mg, 0.59
mmol), 10 (obtained from 9 (150 mg), and SOCl (0.5 mL)) to
29 6 4 3
H N F O PS:
[
PdCl
added slowly to the solution of 6 (0.582 g, 1 mmol) also
dissolved in CH Cl (50 mL). The mixture was stirred at room
2 2 2 2
(PhCN) ] (0.380 g, 1 mmol), in CH Cl (50 mL) was
R
2
2
temperature for 15-30 min, and the solvent was partially
evaporated in vacuo. The concentrated solution was treated
with dry n-hexane. The resulting precipitate was filtered,
washed (3 × 30 mL) with n-hexane, dried, and redissolved in
hot alcohol layered with n-hexane. Slow evaporation gave 7
as orange brown crystals (0.721 g), 75%. Mp: 175 °C dec; IR:
3
2
give the crude product 12. Purification on a silica gel column
by eluting with 30% ethyl acetate and hexane yielded 12 (108
mg) in 78% yield. Mp: 55-56 °C; IR: 3330, 3246, 3055, 2921,
-
1
1
2851, 1682, 1647, 1520, 1344, 1104, 977, 724, 689 cm ; H
NMR: δ 1.10-1.50 (m, 5H), 1.55-1.80 (m, 3H), 1.99-2.05 (m,
2H), 2.83 (d, J ) 10.83 Hz, 3H), 3.12 (d, J ) 11.28 Hz, 3H),
3.60 (m, 1H), 3.78 (d, J ) 12.26 Hz, 1H), 4.04 (d, J ) 12.29
Hz, 1H), 4.07 (m, 1H), 4.15 (br, 1H), 7.33-7.55 (m, 5H), 7.63
2
3
4
8
128, 1532, 1685, 1462, 1277, 800, 792, 598 cm-1; 1H NMR: δ
.22 (2 overlapping doublets, 6H), 4.12 (d, J ) 13.4 Hz, 1H),
3
1
.85 (d, J )13.5 Hz, 1H), 7.49-8.79 (m, 10H); P NMR: δ
1
9
8.2; F NMR δ -140.5 (m, 2F), -151.4 (m, 2F). Anal. Calcd
PSPdCl C, 34.83; H, 2.53; N, 14.78. Found:
C, 34.49, H, 2.54, N, 14.45.
3
1
19
for C22
H
19
N
8
F
4
O
3
2
(s, 1H), 7.91 (s, 1H), 7.95-8.16 (m, 3H); P NMR: δ 83.2;
F
NMR: δ -143.8 (m, 2F), -161.0 (m, 2F); 13C NMR: δ 158.4,
147.7, 146.6 (m), 142.7 (m), 138.9, 138.1 (m), 134.9, 134.4 (m),
Syn th esis of Meth yl 4-(Cycloh exyla m in o)-2,3,5,6-tet-
r a flu or oben zoa te (8). A solution of methyl 4-azido-2,3,5,6-
tetrafluorobenzoate (0.124 g, 0.5 mmol) in cyclohexane (100
mL) was photolyzed for 4 h. The crude photolysis products
3
1
13
31
13
132.4 [J ( P- C), 10.7 Hz], 131.8 [J ( P- C), 2.5 Hz], 130.9,
1
3
19
128.8, 127.7, 127.5, 123.6, 121.9, 53.3 [CH, J ( C- F), 4.6 Hz],
3
1
13
31
13
51.7, 38.3 [J ( P- C), 6.2 Hz], 34.3 [J ( P- C), 12.2 Hz], 34.0,

Chemistry of Bifunctional Photoprobes
5.1, 24.4. Anal. Calcd for C28 PS: C, 52.66; H, 4.89;
J . Org. Chem., Vol. 62, No. 9, 1997 2807
2
H
31
N
6
F
4
O
3
mixture was incubated for 30 min before extraction into 1 ml
N, 13.16. Found: C, 52.95; H, 4.75; N, 12.96. HPLC t
R
4.9
of CHCl
3 2
. This solution was dried under a stream of dry N ,
min.
washed three times with 1 mL of saline to remove unreacted
PdCl4 , and redissolved in saline (1 mL). Radiochemical
1
09
2-
Syn th esis of Cycloh exa n e Ad d u ct 13. A solution of
PdCl (PhCN) ] (0.380 g, 1 mmol) in CH Cl (30 mL) was added
slowly to the solution of 12 (0.638 g, 1 mmol) also dissolved in
CH Cl (30 mL). The mixture was stirred at room temperature
[
2
2
2
2
purity was determined by spotting 5 µL of the complex solution
on paper chromatography strips (Gelman Pads 10 cm × 0.7
cm) and developing with acetone, ethyl acetate, and saline.
After these solvents reached the top of the strip, the strips
were removed and air dried. Each strip was cut into three
equal sections and counted using a NaI(Tl) counter (Ortec
2
2
for 15-30 min, and the solvent was partially evaporated in
vacuo. The concentrated solution was treated with dry n-
hexane, and the precipitate was filtered and washed (3 × 30
mL) with n-hexane, dried, dissolved in hot acetonitrile, and
slowly evaporated to give orange brown crystals (0.711 g) in
109
2-
4
system). The same solvent systems were used for the PdCl
stock solutions as a control. Radiographic scans were per-
formed on a Bioscan (system 200 imaging scanner) radioana-
8
1
7% yield. Mp: 225 °C dec; IR: 3395, 2921, 2867, 1690, 1521,
361, 1321, 987, 797, 728, 596 cm- ; H NMR: δ 1.08-1.34
1
1
1
09
lytic system. In the isolated complex, the
migrated with acetone (R ) 9.0, 98.1 ( 1.2%) and ethyl acetate
) 9.0, 99 ( 0.5%), but remains at the origin with saline
Pd activity
(
m, 5H), 1.55-1.80 (m, 3H), 1.95 (m, 2H), 3.36 (d, J ) 8.06
f
Hz, 3H), 3.54 (d, J ) 9.73 Hz, 3H), 3.62 (m, 1H), 3.95 (br s,
(
(R
R
f
f
1
1
δ
H), 4.21 (d, J ) 12.83 Hz, 1H), 4.78 (br, 1H), 4.87 (d, J )
31 19
) 2.0, 99 ( 1.3%). In contrast the 1 Pd activity in PdCl
09
2-
4
3.47 Hz, 1H), 7.4-8.1 (m, 10H); P NMR: δ 91.2; F NMR:
-143.6 (m, 2F), -160.6 (m, 2F). Anal. Calcd for
showed exactly the opposite behavior, staying at the origin for
acetone (R ) 2.0, 91 ( 0.3%) and ethyl acetate (R ) 2.0, 98
1.3%), but moving with saline (R ) 9.0, 95 ( 2.1%).
C
28
H
31
N
6
F
4
O
3
PSPdCl
2
: C, 41.22; H, 3.83; N, 10.30. Found:
1.9 min.
f
f
(
f
C, 41.38; H, 3.71; N, 10.42. HPLC t
Syn th esis of th e P d Com p lex.
R
1
09
109
Pd was obtained from
a Pd metal target irradiated at the Missouri University
Ack n ow led gm en t. The work was supported by
Research Reactor (MURR) for 15 min in a thermal flux of 8 ×
funds provided by DOE Grant DEFG0289ER 60875.
NMR data were collected on instruments purchased
with funds from NSF grants 8908304 and 9221835. We
appreciate the help of Dr. Guo and Mr. Gao of the NMR
facility for help with the decoupling experiments. The
authors thank Dr. Silvia J urisson for providing access
to the NaI counter and the Bioscan analysis system.
1
3
2
-1
109
1
0
neutrons/cm s . The entire Pd sample containing Pd
(
10 mCi/mg) was dissolved in aqua regia and evaporated to
2
-
dryness in a boiling water bath. The resulting PdCl
4
and
1
09
2-
4
PdCl
was dissolved in 1 mL of saline and evaporated to
dryness. The sample was diluted with saline (1 mL) and
adjusted to pH 5 by NaOH.
The 109Pd complex of 6 was prepared by slowly adding an
ethanolic solution of 6 (15 mg in 1 mL, 25.8 µmol) to
1
09
2-
4
approximately 7 mCi of
PdCl
in 1 mL of saline. The
J O961867B