Oxalate/Hydrogen Peroxide Chemiluminescence Reaction. A I9F NMR Probe of the Reaction Mechanism Hitesh P. Chokshi, Michael Barbush, Robert G . Carlson,? Richard S. Givens,? Ted Kuwana and Richard L. Schowen The Center for Bioanalytical Research and Department of Chemistry, Malott Hall, The University of Kansas, Lawrence, KS 66045-0046, USA
The mechanism of the oxalate/hydrogen peroxide chemiluminescence reaction has been examined by magnetic resonance techniques. Investigation of the reactive intermediates involved in chemiluminescence was carried out with bis(2,6-difluorophenyl)oxalate (DFPO) using 19F NMR to probe its reactions with aqueous hydrogen peroxide. Formation and reactions of the intermediate hydroperoxy oxalate ester B along with the formation of the half ester product C and difluorophenol D were monitored by 19F NMR. When the reaction of DFPO and aqueous hydrogen peroxide was carried out in the presence of dansylphenylalanine, a typical fluorescent analyte, the intensity of the resonance due to the intermediate B was diminished in direct proportion to the concentration of the analyte. Comparison of the timelintensity profile of the chemiluminescence emission with that of the I9F NMR transient suggests that the hydroperoxy oxalate ester B is the likely ‘reactive’ intermediate, capable of participating in a chemically initiated electron exchange luminescence mechanism.
as a single entity, a situation we have termed ‘pooled intermediates’ (Orlovic et al., 1989).
INTRODUCTION The reaction of diaryl oxalates with hydrogen peroxide in the presence of certain fluorophores produces light in addition to the oxidation and hydrolysis products of the ester. For almost two decades the mechanism of this chemiluminescence reaction has been of interest to chemists because of its complex nature and recently because of its application to chemical analysis at trace levels. The mechanism involves highly reactive intermediates which are alleged to activate the fluorophore through electron transfer processes (McCapra et al., 1981; Schuster and Schmidt, 1982; Givens and Schowen, 1989; Givens et al., 1989). Insight into the nature of these reactive intermediates will be of great utility in designing new oxalates and oxalate-like compounds, and in selecting optimum parameters for the chemiluminescence-based detection for trace assays (Givens el a!., 1989). Rauhut (1969) and McCapra (1968) proposed dioxetanedione as the reactive intermediate which acted as the electron acceptor. The intermediacy of dioxetanedione has never received direct confirmation. However, kinetic studies by Catherall et al. (1984) indicated that one aryl group of the diaryl ester was still present at the reactive intermediate stage, thus establishing that the dioxetanedione was not a crucial reactive intermediate. Our recent studies have confirmed Catherall’s observation and have shown that several reactive intermediates may contribute to the chemiluminescence process (Alvarez et al., 1986). Under certain conditions, notably in aqueous binary solvents which are often encountered in reverse phase HPLC separations, the intermediates can be treated kinetically t Authors
to whom correspondence should be addressed.
O w 0 ,
I
0-0 Dioxetanedione
We have sought information on the structure of the reactive intermediates from several perspectives. In this paper, we present preliminary results of the I9F NMR probe of the oxalate hydrogen peroxide chemiluminescence reaction. NMR studies together with our recent kinetic studies (Orlovic et al., 1989) have provided the first direct structural evidence of the intermediates in the chemiluminescence experiments. ~
~
~~
EXPERIMENTAL 19
F NMR experiments were carried out either on a Varian X L 300 (Sunnyvale, CA, USA) or a Bruker AM 500 (Billerica, MA, USA) spectrometer and the chemical shifts are reported in parts per million (6) relative to a,a,a-trifluorotoluene (6 = -63.7 ppm). NMR experiments were done with a 5 mm probe and the solvent composition of the final solution was 75% [ DJacetonitrile and 25% water by volume. Hydrogen peroxide was added last and the NMR acquisition was started within approximately 200 s. An array of NMR experiments was set up so that each spectrum was obtained with the same number of scans (16 scans, 40 s), and the data collection was continued for 1200 or more seconds. Each spectrum was then plotted and the peak heights were measured to calculate the percent for the reactant, the intermediate, and the products. DFPO was synthesized in our laboratories and its purity was checked by ‘H, I3C, I9F NMR, infrared ( I R ) and mass spectroscopy. The stock solution of DFPO was prepared in
CCC-0269-3879/90/0096-0099$2.00
96 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO 3, 1990
@ Heyden & Son Limited, 1990
NMR PROBE OF OXALATEIHYDROGEN PEROXIDE
RESULTS AND DISCUSSION
Figure 1. ''F resonances of DFPO (A), intermediate B, half ester C and DFP (D) monitored for reaction of [DFPO]=51 RIM and [H202]= 240 mM in 75% aqueous CD,CN.
[DJacetonitrile (99 atom % D) which was purchased from Aldrich Chemical Co (Millwaukee, WI, USA). Hydrogen peroxide (30%) was obtained from Fisher Scientific H325 (West Haven, CT, USA) and its stock solution was prepared in water. Dansylphenylalanine was obtained from Sigma (St Louis, MO, USA) and its stock solution was prepared in [DJacetonitrile. 19 F N M R experiments were carried out using the following conditions: [DFPO] = 51 mM; [DNS-Phe] = 0-22 mM; [H,O,] = 240 mM; 75% CD3CN+25% water at 25 "C.
To monitor the chemiluminescence reaction by ''F NMR, bis(2,6-difluorophenyI) oxalate (DFPO) was chosen. The I9F N M R signals for DFPO ( A ) and difluorophenol (DFP, D) appeared at -127.5 and -135.6 ppm, respectively. When DFPO was treated with water alone, resonances at -128.2 and -135.6ppm (DFP) appeared and increased steadily whereas the DFPO resonance at -127.5 ppm decreased. The signal at -128.2 ppm was assigned to the half ester C based upon the following evidence. The acid chloride E, prepared by a procedure similar to tha: used for other haloaryl oxalates (Baker and Schumacher, 1964), showed a resonance at -127.2 ppm. On addition of water to the acid chloride E, a resonance due to the half ester C appeared at -128.2 ppm. When DFPO was treated with aqueous hydrogen peroxide, an additional resonance appeared at - 127.7 ppm (Fig. l ) , which increased and then slowly decayed, was assigned to the hydroperoxy oxalate ester B from the
Table 1. Structures and I9F NMR chemical shifts for oxalates and corresponding benzoic acid derivatives Benzoic Acid Series
Oxalate series
A, DFPO
0-C-C-0-0-H
w
--127.7
-111.9 \
F
F
2.6-Difluoroperbenzoic acid
B, Hydroperoxy oxalate ester
-128.2
-1 13.2
/ \ \
F
C. Half ester
F
2.6-Difluorobenzoic acid
-135.6
F D, DFP F
- 1 27.2
F E, Acid chloride
@ Heyden & Son Limited, 1990
-111.7
2,6-Difluorobenzoyl chloride
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO 3. 1990 97
P. CHOKSHI E T A L
A
('
I1
I
I
80
A
-
C(Ha1Tesier)
;.........: 1
xin 730
Figure 2. 'F resonances of DFPO (A), intermediate B, half ester C and DFP (D) monitored for reaction of [DFPO]=51 mM, [H202]= 240 mM and (DNS-Phe]=8 mM in 75% aqueous CDJN.
following evidence. The chemical shifts determined for the oxalate esters (Table 1) were compared with those observed for 2,6-difluorobenzoyl chloride (-1 11.7 ppm), 2,6-difluoroperbenzoic acid (-1 11.9 ppm) and 2,6difluorobenzoic acid (-1 13.2 ppm); the latter two were generated independently by reaction of the acid chloride with water or aqueous hydrogen peroxide. The direction and magnitude of changes in chemical shifts in the series suggested that in the oxalate series the - 127.7 ppm signal could be assigned to intermediate B. The low intensity of the signal at -127.7 ppm suggested that the concentration of the intermediate was very low. On the other hand, when the same reaction was carried out in the EtOAc+ triethylamine system (Alvarez et aZ., 1986), the intensity of the -127.7 ppm resonance indicated a substantially greater concentration of the intermediate. When DFPO was treated with aqueous hydrogen peroxide in the presence of dansyl phenylalanine (DNSPhe) as a fluorophore, conditions suitable for chemiluminescence emission from DNS-Phe, the intensity of the resonance at -127.7 ppm was diminished in direct proportion to the DNS-Phe concentration (Fig. 2). This behavior was clearly demonstrated in plots of the rise and fall of DFPO, B, C , and D in the experiments without DNS-Phe (Fig. 3), at low concentrations of DNS-Phe (Fig. 4) and at high concentrations of DNSPhe. The resonance due to the intermediate B was not observed in experiments at high concentration of DNSPhe. A comparative plot for the rise and fall of the intermediate B in these experiments is shown in Fig. 5. Parallel studies under identical conditions, but monitoring the chemiluminescence emission, showed a biex-
'5 60 3:
A
I
Y
2
40
8 20
0 0
1000 Time@=)
2000
Figure 3. Rise and fall of DFPO (A), intermediate b, half ester C and DFP (D) for reaction of [DFPO]=51 mM and [H,O2]=240m~.
98 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990
A(DFPO) B(lnlermedlar)
O
0
I000
2000
Time(sec) Figure 4. Rise and fall of DFPO (A), intermediate B, half ester C and DFP ( 0 )for reaction of [DFPO]=51 m M , [ H 2 0 2 ] = 2 4 0 m ~ and [DNS-Phe] =8 mM.
O
0
[DNS-Phe)=Om M [DNS-Phel=8 m M
1000 Time(sec)
2000
Figure 5. Enlargement of plots of the rise and fall of intermediate B for reaction of [DFPO] =51 mM and [H202]=240 mM with DNS-Phe concentration a t 0 mM [0) and 8 mM ( X ) from Figs 3 and 4.
ponential timelintensity curve with a same characteristic time to reach a maximum (T,,, = 8 min) as observed for the intermediate B in the NMR experiment. To date, there is no evidence for a second intermediate at concentrations detectable with 19FNMR. The similarity of the kinetic profile from the NMR transient and the time/ intensity signal for the chemiluminescence reaction strongly suggest a common peroxyoxalate intermediate, which we proposed to be B. The results further suggest that intermediate B serves as an electron acceptor and is capable of participating in a chemically initiated electron exchange luminescence (CIEEL) mechanism (McCapra et al., 1981; Schuster and Schmidt, 1982; Givens and Schowen, 1989; Givens et al., 1989). Our results do not rule out the subsequent formation of a short-lived proton or valence tautomeric form of B. However, the likelihood that such a short-lived, low concentration intermediqe could take part in a bimolecular electron transfer process woyld appear to be exceedingly improbable. Thus, the favored candidate for the 'active' intermediate is the hydroperoxy aryl oxalate B. Acknowledgements We thank the Kansas Technology Enterprise Commission, Oread Laboratories, Inc. and NSF (Grant No: MACRO-ROA CHE 884263) for support of this research.
@ Heyden & Son Limited, 1990
N M R PROBE OF OXALATEIH YD R OGEN PEROXIDE
REFERENCES Alvarez, F. J., Parekh, N.J., Matuszewski. B..Givens, R. S., Higuchi, T. and Schowen. R. L. (1986). J. Am. Chem. SOC.108, 6435. Baker, J. W. and Schumacher, 1. (1964). J. Chem. and Eng. Data. 584. Catherall, C. L. R., Palmer, T. F. and Cundall, R. B. (1984). J. Chem. SOC.,Faraday Trans. 2, 80, 823, 837. Givens. R. S. and Schowen, R. L. (1989). In Chemiluminescenceand Photochemical Reaction Detection in Chromatography, ed. by J. W. Birks, Ch. 5, p p 125147. VCH, Inc., New York. Givens, R. S., Schowen, R. L., Stobaugh, J., Alvarez, F., Parekh, N., Matuszewski, B., Kawasaki, T., Wong, 0.. Chokshi, H., Orlovic, M. and Nakashima. K. (1989). In Luminescence Applications in Biological, Chemical, Environmental and Hydrological Sciences, ed. by M. C. Goldberg, Ch 8, pp 127-154, ACS Symposium Series 383.
@ Heyden & Son Limited, 1990
McCapra, F., Perring, K., Hart, R. J. and Hann, R. A. (1981). Tetrahedron Lett. 5087. McCapra, F. (1968). Chem. Cornrnun. 155. Orlovic, M., Schowen, R. L., Givens, R. S., Alvarez, F., Matuszewski, B. and Parekh, N. (1989). J. Org. Chem. 54, 3605. Rauhut, M. M. (1969). Acc. Chem. Res. 2. 80. Schuster, G. B. and Schmidt, S . P. Adv. Phys. Org. Chem. (1982). 18, 187-238.
Received 21 September 1989; accepted 9 October 1989.
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990 99
The mechanism of the oxalate/hydrogen peroxide chemiluminescence reaction has been examined by magnetic resonance techniques. Investigation of the reactive intermediates involved in chemiluminescence was carried out with bis(2,6-difluorophenyl)oxalate (DFPO) using 19F NMR to probe its reactions with aqueous hydrogen peroxide. Formation and reactions of the intermediate hydroperoxy oxalate ester B along with the formation of the half ester product C and difluorophenol D were monitored by 19F NMR. When the reaction of DFPO and aqueous hydrogen peroxide was carried out in the presence of dansylphenylalanine, a typical fluorescent analyte, the intensity of the resonance due to the intermediate B was diminished in direct proportion to the concentration of the analyte. Comparison of the timelintensity profile of the chemiluminescence emission with that of the I9F NMR transient suggests that the hydroperoxy oxalate ester B is the likely ‘reactive’ intermediate, capable of participating in a chemically initiated electron exchange luminescence mechanism.
as a single entity, a situation we have termed ‘pooled intermediates’ (Orlovic et al., 1989).
INTRODUCTION The reaction of diaryl oxalates with hydrogen peroxide in the presence of certain fluorophores produces light in addition to the oxidation and hydrolysis products of the ester. For almost two decades the mechanism of this chemiluminescence reaction has been of interest to chemists because of its complex nature and recently because of its application to chemical analysis at trace levels. The mechanism involves highly reactive intermediates which are alleged to activate the fluorophore through electron transfer processes (McCapra et al., 1981; Schuster and Schmidt, 1982; Givens and Schowen, 1989; Givens et al., 1989). Insight into the nature of these reactive intermediates will be of great utility in designing new oxalates and oxalate-like compounds, and in selecting optimum parameters for the chemiluminescence-based detection for trace assays (Givens el a!., 1989). Rauhut (1969) and McCapra (1968) proposed dioxetanedione as the reactive intermediate which acted as the electron acceptor. The intermediacy of dioxetanedione has never received direct confirmation. However, kinetic studies by Catherall et al. (1984) indicated that one aryl group of the diaryl ester was still present at the reactive intermediate stage, thus establishing that the dioxetanedione was not a crucial reactive intermediate. Our recent studies have confirmed Catherall’s observation and have shown that several reactive intermediates may contribute to the chemiluminescence process (Alvarez et al., 1986). Under certain conditions, notably in aqueous binary solvents which are often encountered in reverse phase HPLC separations, the intermediates can be treated kinetically t Authors
to whom correspondence should be addressed.
O w 0 ,
I
0-0 Dioxetanedione
We have sought information on the structure of the reactive intermediates from several perspectives. In this paper, we present preliminary results of the I9F NMR probe of the oxalate hydrogen peroxide chemiluminescence reaction. NMR studies together with our recent kinetic studies (Orlovic et al., 1989) have provided the first direct structural evidence of the intermediates in the chemiluminescence experiments. ~
~
~~
EXPERIMENTAL 19
F NMR experiments were carried out either on a Varian X L 300 (Sunnyvale, CA, USA) or a Bruker AM 500 (Billerica, MA, USA) spectrometer and the chemical shifts are reported in parts per million (6) relative to a,a,a-trifluorotoluene (6 = -63.7 ppm). NMR experiments were done with a 5 mm probe and the solvent composition of the final solution was 75% [ DJacetonitrile and 25% water by volume. Hydrogen peroxide was added last and the NMR acquisition was started within approximately 200 s. An array of NMR experiments was set up so that each spectrum was obtained with the same number of scans (16 scans, 40 s), and the data collection was continued for 1200 or more seconds. Each spectrum was then plotted and the peak heights were measured to calculate the percent for the reactant, the intermediate, and the products. DFPO was synthesized in our laboratories and its purity was checked by ‘H, I3C, I9F NMR, infrared ( I R ) and mass spectroscopy. The stock solution of DFPO was prepared in
CCC-0269-3879/90/0096-0099$2.00
96 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO 3, 1990
@ Heyden & Son Limited, 1990
NMR PROBE OF OXALATEIHYDROGEN PEROXIDE
RESULTS AND DISCUSSION
Figure 1. ''F resonances of DFPO (A), intermediate B, half ester C and DFP (D) monitored for reaction of [DFPO]=51 RIM and [H202]= 240 mM in 75% aqueous CD,CN.
[DJacetonitrile (99 atom % D) which was purchased from Aldrich Chemical Co (Millwaukee, WI, USA). Hydrogen peroxide (30%) was obtained from Fisher Scientific H325 (West Haven, CT, USA) and its stock solution was prepared in water. Dansylphenylalanine was obtained from Sigma (St Louis, MO, USA) and its stock solution was prepared in [DJacetonitrile. 19 F N M R experiments were carried out using the following conditions: [DFPO] = 51 mM; [DNS-Phe] = 0-22 mM; [H,O,] = 240 mM; 75% CD3CN+25% water at 25 "C.
To monitor the chemiluminescence reaction by ''F NMR, bis(2,6-difluorophenyI) oxalate (DFPO) was chosen. The I9F N M R signals for DFPO ( A ) and difluorophenol (DFP, D) appeared at -127.5 and -135.6 ppm, respectively. When DFPO was treated with water alone, resonances at -128.2 and -135.6ppm (DFP) appeared and increased steadily whereas the DFPO resonance at -127.5 ppm decreased. The signal at -128.2 ppm was assigned to the half ester C based upon the following evidence. The acid chloride E, prepared by a procedure similar to tha: used for other haloaryl oxalates (Baker and Schumacher, 1964), showed a resonance at -127.2 ppm. On addition of water to the acid chloride E, a resonance due to the half ester C appeared at -128.2 ppm. When DFPO was treated with aqueous hydrogen peroxide, an additional resonance appeared at - 127.7 ppm (Fig. l ) , which increased and then slowly decayed, was assigned to the hydroperoxy oxalate ester B from the
Table 1. Structures and I9F NMR chemical shifts for oxalates and corresponding benzoic acid derivatives Benzoic Acid Series
Oxalate series
A, DFPO
0-C-C-0-0-H
w
--127.7
-111.9 \
F
F
2.6-Difluoroperbenzoic acid
B, Hydroperoxy oxalate ester
-128.2
-1 13.2
/ \ \
F
C. Half ester
F
2.6-Difluorobenzoic acid
-135.6
F D, DFP F
- 1 27.2
F E, Acid chloride
@ Heyden & Son Limited, 1990
-111.7
2,6-Difluorobenzoyl chloride
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO 3. 1990 97
P. CHOKSHI E T A L
A
('
I1
I
I
80
A
-
C(Ha1Tesier)
;.........: 1
xin 730
Figure 2. 'F resonances of DFPO (A), intermediate B, half ester C and DFP (D) monitored for reaction of [DFPO]=51 mM, [H202]= 240 mM and (DNS-Phe]=8 mM in 75% aqueous CDJN.
following evidence. The chemical shifts determined for the oxalate esters (Table 1) were compared with those observed for 2,6-difluorobenzoyl chloride (-1 11.7 ppm), 2,6-difluoroperbenzoic acid (-1 11.9 ppm) and 2,6difluorobenzoic acid (-1 13.2 ppm); the latter two were generated independently by reaction of the acid chloride with water or aqueous hydrogen peroxide. The direction and magnitude of changes in chemical shifts in the series suggested that in the oxalate series the - 127.7 ppm signal could be assigned to intermediate B. The low intensity of the signal at -127.7 ppm suggested that the concentration of the intermediate was very low. On the other hand, when the same reaction was carried out in the EtOAc+ triethylamine system (Alvarez et aZ., 1986), the intensity of the -127.7 ppm resonance indicated a substantially greater concentration of the intermediate. When DFPO was treated with aqueous hydrogen peroxide in the presence of dansyl phenylalanine (DNSPhe) as a fluorophore, conditions suitable for chemiluminescence emission from DNS-Phe, the intensity of the resonance at -127.7 ppm was diminished in direct proportion to the DNS-Phe concentration (Fig. 2). This behavior was clearly demonstrated in plots of the rise and fall of DFPO, B, C , and D in the experiments without DNS-Phe (Fig. 3), at low concentrations of DNS-Phe (Fig. 4) and at high concentrations of DNSPhe. The resonance due to the intermediate B was not observed in experiments at high concentration of DNSPhe. A comparative plot for the rise and fall of the intermediate B in these experiments is shown in Fig. 5. Parallel studies under identical conditions, but monitoring the chemiluminescence emission, showed a biex-
'5 60 3:
A
I
Y
2
40
8 20
0 0
1000 Time@=)
2000
Figure 3. Rise and fall of DFPO (A), intermediate b, half ester C and DFP (D) for reaction of [DFPO]=51 mM and [H,O2]=240m~.
98 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990
A(DFPO) B(lnlermedlar)
O
0
I000
2000
Time(sec) Figure 4. Rise and fall of DFPO (A), intermediate B, half ester C and DFP ( 0 )for reaction of [DFPO]=51 m M , [ H 2 0 2 ] = 2 4 0 m ~ and [DNS-Phe] =8 mM.
O
0
[DNS-Phe)=Om M [DNS-Phel=8 m M
1000 Time(sec)
2000
Figure 5. Enlargement of plots of the rise and fall of intermediate B for reaction of [DFPO] =51 mM and [H202]=240 mM with DNS-Phe concentration a t 0 mM [0) and 8 mM ( X ) from Figs 3 and 4.
ponential timelintensity curve with a same characteristic time to reach a maximum (T,,, = 8 min) as observed for the intermediate B in the NMR experiment. To date, there is no evidence for a second intermediate at concentrations detectable with 19FNMR. The similarity of the kinetic profile from the NMR transient and the time/ intensity signal for the chemiluminescence reaction strongly suggest a common peroxyoxalate intermediate, which we proposed to be B. The results further suggest that intermediate B serves as an electron acceptor and is capable of participating in a chemically initiated electron exchange luminescence (CIEEL) mechanism (McCapra et al., 1981; Schuster and Schmidt, 1982; Givens and Schowen, 1989; Givens et al., 1989). Our results do not rule out the subsequent formation of a short-lived proton or valence tautomeric form of B. However, the likelihood that such a short-lived, low concentration intermediqe could take part in a bimolecular electron transfer process woyld appear to be exceedingly improbable. Thus, the favored candidate for the 'active' intermediate is the hydroperoxy aryl oxalate B. Acknowledgements We thank the Kansas Technology Enterprise Commission, Oread Laboratories, Inc. and NSF (Grant No: MACRO-ROA CHE 884263) for support of this research.
@ Heyden & Son Limited, 1990
N M R PROBE OF OXALATEIH YD R OGEN PEROXIDE
REFERENCES Alvarez, F. J., Parekh, N.J., Matuszewski. B..Givens, R. S., Higuchi, T. and Schowen. R. L. (1986). J. Am. Chem. SOC.108, 6435. Baker, J. W. and Schumacher, 1. (1964). J. Chem. and Eng. Data. 584. Catherall, C. L. R., Palmer, T. F. and Cundall, R. B. (1984). J. Chem. SOC.,Faraday Trans. 2, 80, 823, 837. Givens. R. S. and Schowen, R. L. (1989). In Chemiluminescenceand Photochemical Reaction Detection in Chromatography, ed. by J. W. Birks, Ch. 5, p p 125147. VCH, Inc., New York. Givens, R. S., Schowen, R. L., Stobaugh, J., Alvarez, F., Parekh, N., Matuszewski, B., Kawasaki, T., Wong, 0.. Chokshi, H., Orlovic, M. and Nakashima. K. (1989). In Luminescence Applications in Biological, Chemical, Environmental and Hydrological Sciences, ed. by M. C. Goldberg, Ch 8, pp 127-154, ACS Symposium Series 383.
@ Heyden & Son Limited, 1990
McCapra, F., Perring, K., Hart, R. J. and Hann, R. A. (1981). Tetrahedron Lett. 5087. McCapra, F. (1968). Chem. Cornrnun. 155. Orlovic, M., Schowen, R. L., Givens, R. S., Alvarez, F., Matuszewski, B. and Parekh, N. (1989). J. Org. Chem. 54, 3605. Rauhut, M. M. (1969). Acc. Chem. Res. 2. 80. Schuster, G. B. and Schmidt, S . P. Adv. Phys. Org. Chem. (1982). 18, 187-238.
Received 21 September 1989; accepted 9 October 1989.
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990 99
