Formation and verification of the structure of the 1-fluorenylmethyl chloroformative derivative of sulfamethazine

Article abstract of DOI:10.1021/ac9505280

Sulfamethazine (SMZ) is derivatized with 1-fluorenylmethyl chloroformate (FMOC) to form the fluorescent adduct SMZ-FMOC. Conditions for formation are optimized with respect to pH, reagent concentration, and reagent ratio. Reagent and product profiles (including the hydrolysis byproduct FMOC-OH) versus time are followed by reversed phase HPLC with UV absorbance detection. FMOC-SMZ has been crystallized, its composition confirmed by microanalysis, and its structure corroborated by IR and NMR spectroscopy. From 10 down to 1 ppm, there is clear gentle curvature in the fluorescence intensity of SMZ-FMOC. The linear response range extends from above 100 ppb down to about 100 ppt, and an increase in sensitivity for the fluorescent detection of FMOC-SMZ (over the usual UV absorbance detection of SMZ) is calculated to be better than 3 orders of magnitude.

Full text of DOI:10.1021/ac9505280

Anal. Chem. 1996, 68, 86-92
Formation and Verification of the Structure of the
1-Fluorenylmethyl Chloroformate Derivative of
Sulfamethazine
Gao Shan (Sam) Liang, Zane (Zhi Yu) Zhang, Warren L. Baker, and Reginald F. Cross*
School of Chemical Sciences, Swinburne University of Technology, John Street, Hawthorn, Victoria 3122, Australia
analytical effort in SFA residue analysis worldwide. Bovine milk
in the USA,10 salmon flesh in Canada,11 beef, pork, chicken, ham,
sausage, bacon, roast beef,12,13 and milk14 in Japan, and pork in
the Netherlands15 have been the subjects of SFA analysis in recent
years. However, the dependence of sectors of the meat producing
industry on the SFAs16 indicates that the SFAs will continue in
their prophylactic veterinary role. Hence, there is the continued
need for sensitive methods of analysis of these antiinfective drugs.
Most instrumental methods of separation have been tried for
the SFA,17 but specific and quantitative methods are limited to
GC,18-22 GC/ MS,7,22 HPLC (for some recent examples, see refs
17, 18, 23-32), thermospray LC/ MS,33-35 CZE,36 MECC,37 and
CZE-MS.34,38 HPLC has been the preferred basis for analysis.
Sulfamethazine (SMZ) is derivatized with 1 -fluorenyl-
methyl chloroformate (FMOC) to form the fluorescent
adduct SMZ-FMOC. Conditions for formation are opti-
mized with respect to pH, reagent concentration, and
reagent ratio. Reagent and product profiles (including the
hydrolysis byproduct FMOC-OH) versus time are followed
by reversed phase HP LC with UV absorbance detection.
FMOC-SMZ has been crystallized, its composition con-
firmed by microanalysis, and its structure corroborated
by IR and NMR spectroscopy. From 1 0 down to 1 ppm,
there is clear gentle curvature in the fluorescence intensity
of SMZ-FMOC. The linear response range extends from
above 1 0 0 ppb down to about 1 0 0 ppt, and an increase
in sensitivity for the fluorescent detection of FMOC-SMZ
(over the usual UV absorbance detection of SMZ) is
calculated to be better than 3 orders of magnitude.
(9) Cronic Toxicity and Carcinogonicity of Sulphamethazine in B6CF1 Mice.
Technical report for experiment no. 418, National Center for Toxilogical
Research, Jefferson, AR; Fed. Regist. 1 9 8 8 , 53, FR9492.
(10) Smedley, M. D.; Weber, J. D. J. Assoc. Off. Anal. Chem. 1 9 9 0 , 73, 875-9.
(11) Pleasance, S.; Blay, P.; William, M. A. Q.; O’Hara, G. J. Chromatogr. 1 9 9 1 ,
558, 155-73.
(12) Takeda, N.; Akiyama, Y. J. J. Chromatogr. 1 9 9 1 , 558, 175-80.
(13) Kishihara, S.; Shimokawa, C.; Izumi, H. Ishikawa-ken Eisei Kogai Kenkyusho
Nenpo 1 9 8 3 , 198-204.
(14) Takeda, N.; Akiyama, Y. J. J. Chromatogr. 1 9 9 2 , 607, 31-5.
(15) Ackermans, M. T.; Beckers, J. L.; Everaerts, F. M.; Hoogland, H.; Tomassen,
M. J. H. J. Chromatogr. 1 9 9 2 , 596, 101-9.
(16) Australian Pig Industry Policy Council, Annual Report, 1991-2.
(17) Ricci, M. C.; Cross, R. F. J. Microcolumn Sep. 1 9 9 3 , 5, 207-15.
(18) Duan, R. J. J. Assoc. Off. Anal. Chem. 1 9 7 1 , 54, 1277-82.
(19) Goodspeed, D. P.; Simpson, R. M.; Ashworth, R. B.; Shafer, J. W.; Cook, H.
R. J. Assoc. Off. Anal. Chem. 1 9 7 8 , 61, 1050-3.
Sulfonamides (SFAs) are widely used for the prevention of
infectious diseases in livestock animals. However, there is a
consumer expectation of residue-free food, and residues of SFAs
in foods derived from treated animals could lead to the develop-
ment of drug-resistant strains of microorganisms.1 Consequently,
the SFAs have been under surveillance for several years. The
U.S. Department of Agriculture,2,3 the Australian Meat and
Livestock Research and Development Corporation,4-6 the Euro-
pean Economic Community,7 and the Food and Agriculture
Organization of the United Nations8 have all concerned themselves
with veterinary drug residues. The SFAs in general have been
the principal class targeted, and sulfamethazine (SMZ, alternately
called sulfadimidine) has been preeminent due to a study indicat-
ing possible carcinogenicity.9 Hence, there has been a huge
(20) Holder, C. L.; Thompson, H. C., Jr.; Bowman, M. C. J. Chromatogr. Sci. 1981,
19, 625-33.
(21) Holtmannspotter, H.; Thier, H. P. Dtsch. Lebensm. Rundsch. 1 9 8 2 , 78, 347-
50.
(22) Matusik, J.; Sternal, S.; Barnes, J.; Sphon, J. A. J. Assoc. Off. Anal. Chem.
1 9 9 0 , 73, 529-33.
(23) Ikai, Y.; Oka, H.; Kawamura, N.; Hayakawa, J.; Yamada, M.; Makoto, K.-I.;
Suzuki, H.; Nakazawa, H. J. Chromatogr. 1 9 9 1 , 541, 393-400.
(24) Horie, M.; Saito, K.; Hoshino, Y.; Nose, N.; Hamada, N.; Nakazawa, H. J.
Chromatogr. 1 9 9 0 , 502, 371-8.
(25) Cox, B. L.; Krzeminski, L. F. J. Assoc. Off. Anal. Chem. 1 9 8 2 , 65, 1311-5.
(26) Long, A. R.; Hsieh, L. C.; Malbrough, M. S.; Short, C. R.; Barker, S. A. J.
Agric. Food Chem. 1 9 9 0 , 38, 425-8.
(1) Pratt, W. B.; Fekety, R. The Antimicrobial Drugs; Oxford University Press:
Oxford, UK, 1986.
(2) Compound Evaluation and Analytical Capability; 1985 National Residue
Program Plan; Food Safety and Inspection Service, Science, U.S. Dept. of
Agriculture, US GPO: Washington, DC, 1985.
(3) Guest, G., Director, Center for Veterinary Medicine, USFDA. 47th Annual
Meeting of the Animal Health Institute, Tucson, AZ, May 10-13, 1987.
(4) McCausland, I.; Healey, K.; Wells, R. Emerging Technologies for Testing
Chemical Residues in Meat; AMLRDC Study Team to the USA, 1987.
(5) Pryor, W. J. Study of Important Meat Residue Chemicals. Their Use and
Alternatives; Australian Meat and Live-Stock Research and Development
Corp., Sydney, Australia, 1987.
(6) Pryor, W. J. Potential Meat Contaminants; Australian Meat and Live-Stock
Research and Development Corp., Sydney, Australia, 1988.
(7) Heitzman, R. J., Ed. Veterinary Drug Residues. Residues in food producing
animals and their products. Reference materials and methods; Commission
for the European Communities, EUR 14126 EN, 1992.
(27) Lindsay, H. J.; Smith, P. J.; Love, P. J. Food Analysis Conference, Brisbane,
Australia, 1989.
(28) Aerts, M. M. L.; Beek, W. M. J.; Brinkman, V. A. T. J. Chromatogr. 1 9 8 8 ,
435, 97-112.
(29) Weber, J. D.; Smedley, M. D. J. Assoc. Off. Anal. Chem. 1 9 8 9 , 72, 445-7.
(30) Smedley, M. D.; Weber, J. D. J. Assoc. Off. Anal. Chem. 1 9 9 0 , 73, 875-9.
(31) Long, A. R.; Short, C. R.; Barker, S. A. J. Chromatogr. 1 9 9 0 , 502, 87-9.
(32) Agarwal, V. K. J. Liq. Chromatogr. 1 9 9 0 , 13, 3531-7.
(33) Horie, M.; Saito, K.; Hoshino, Y.; Nose, N.; Tera, M.; Kitsuwa, T.; Nakazawa,
H.; Yamane, Y. Eisei Kagaku 1 9 8 9 , 4, 283-9.
(34) Perkins, J. R.; Parker, C. E.; Tomer, K. B. Am. Soc. Mass Spectrom. 1 9 9 2 ,
(8) Residues of Some Veterinary Drugs in Animals and Foods, 34th Meeting, Joint
FAO/ WHO Expert Committee on Food Additives, Geneva, January 30-
February 8, 1989.
3, 139-49.
(35) Abian, J.; Churchwell, M. I.; Korfmacher, W. I. J. Chromatogr. 1 9 9 3 , 629,
267-76.
86 Analytical Chemistry, Vol. 68, No. 1, January 1, 1996
0003-2700/96/0368-0086$12.00/0 © 1995 American Chemical Society

Figure 1. FMOC reactions with SMZ and water.
Even with derivatization, HPLC, especially reversed phase (RP)
HPLC, does not require as much cleanup as is needed for GC
analyses. Water-based biological systems are inherently compat-
ible with the aqueous mobile phases of RP-HPLC and the gentler
conditions of analysis (lower temperatures) than in GC. The
alternative solution technique of CE is very attractive due to its
higher efficiency, but the inferior concentration sensitivity is
counterproductive in quantitation.
We have examined this derivatization reaction, the optimum
conditions for formation of FMOC-SMZ, and the spectral proper-
ties of the derivative. The proposed structure has been verified.
Concurrent with this study, the fluorescamine derivative has also
been formed and used for the the HPLC analysis of sulfon-
amides.12-14
EXPERIMENTAL SECTION
Chemicals. All solvents used were analytical reagent quality
Most HPLC determinations of the SFA have used UV absorp-
tion detection. However, derivatization with a strongly fluorescent
moiety would clearly enhance detection limits. 9-Fluorenylmethyl
chloroformate (FMOC) was first introduced in HPLC fluorescence
methods based on precolumn derivatization and RP separation of
FMOC-amino acids by Einarsson et al.,39 was later commercialized
by Cunico and others,40 and has found wide application to the
analysis of amino acids and their derivatives.41-46 Under mild
conditions, it reacts with primary and secondary amines and yields
highly fluorescent derivatives. FMOC has also been used for the
derivatization and analyses of other types of amines including
biogenic amines47 and polyamines48-50 and more recentlysduring
the course of this studyshas been used for the analysis of cate-
cholamines51 and amphetamines.52 FMOC should react with the
amine group of SMZ to yield a highly fluorescent FMOC-SMZ
derivative according to the first reaction in Figure 1. (FMOC also
hydrolyzes39,40 according to the second reaction in Figure 1.)
or spectroscopic grade. SMZ was purchased from Sigma Chemi-
cal Co. (St. Louis, MO), and standard solutions of it were prepared
by dissolution in 50% acetone-50% water. FMOC (97%) was
obtained from Fluka Chemika (Buchs, Switzerland) and dissolved
in acetone. The standard HPLC buffer was prepared from NaH2-
PO4 and Milli-Q water and adjusted to pH 3.5 with H3PO4. The
pH 8 reaction buffer was 0.1 M NaHCO3. To study the effect of
pH, reaction solutions were prepared from NaAc-HAc (pH 3-5),
Na2HPO4-NaH2PO4 (pH 6-9), and NaOH (pH 9-10).
Instrumentation. Two HPLC systems were used to monitor
reactions. The first was a conventional system from Waters
(Milford, MA) comprising Model 450 pumps, a 441 fixed-
wavelength (254 nm) UV absorbance detector with a 10 mm path
length flow cell, a 721 data system printer/ plotter, and a 730
gradient controller. A Rheodyne (Cotati, CA) six-port injector with
a 20 µL sample loop was used, and the TMS-capped Micro Park
C18 5 µm, 15 cm × 4 mm i.d. column was supplied by Varian
(Walnut Creek, CA). The stationary phase support was Separa-
tions Group (Hesperia, CA) Vydac IDI TP 80 m2 g-1 silica.
The alternate system was a Varian Vista 5560 HPLC modified
for fast analyses. The conventional flow cell in the Varian UV
200 detector was replaced with the 0.5 µL, 0.5 mm path length
option, and the detector was mounted on the column oven of a
cut-down Perkin-Elmer (Norwalk, CT) LC 65-T UV detector. A
TMS-capped, Varian Micro Pak SP C18 3 µm, 4 cm × 4 mm
column, with PhaseSep 200 m2 g-1 silica, was located directly
below the detector inlet and was connected via 5 cm of 0.13 mm
(0.005 in.) i.d. stainless steel tubing to minimize dispersion. The
Valco (Houston, TX) C14M air-actuated automatic injector was
mounted on the side of the oven and connected through the oven
wall by 10 cm of 0.13 mm i.d. tubing. The wavelength of
maximum absorbance was chosen for detection with this system
(264 nm), and the data system was a Varian CDS 401/ 2.
(36) Ackermans, M. T.; Beckers, J. L.; Everaerts, F. M.; Hoogland, H.; Tomassen,
M. J. H. J. Chromatogr. 1 9 9 2 , 596, 101-9.
(37) Dang, Q. X.; Sun, Z. P.; Ling, D. K. J. Chromatogr. 1 9 9 2 , 603, 259-66.
(38) Pleasance, S.; Thibault, P.; Kelly, J. J. Chromatogr. 1 9 9 2 , 591, 325-39.
(39) Einarsson, S.; Josefsson, B.; Lagerkvist, S. J. Chromatogr. 1 9 8 3 , 282, 609-
18.
(40) Cunico, R. L.; Mayer, A. G.; Dimson, P.; Zimmerman, C. BioChromatography
1 9 8 6 , 1, 6-14.
(41) Naesholm, T.; Sandberg, G.; Ericsson, A. J. Chromatogr. 1 9 8 5 , 396, 225-
36.
(42) Korte, E. GIT-Suppl. 1 9 8 7 , 3, 44-6.
(43) Kisby, G. E.; Roy, D. N.; Spencer, P. S. J. Neurosci. Methods 1 9 8 8 , 26, 45-
54.
(44) Blankenship, D. T.; Krivanek, M. A.; Ackermann, B. L.; Cardin, A. D. Anal.
Biochem. 1 9 8 9 , 178, 227-32.
(45) Teerlink, T.; Boer, D. E. J. Chromatogr. 1 9 8 9 , 491, 418-23.
(46) Guo, Y. J.; Huang, Y. L.; Yao, Z. J. Sepu 1 9 8 9 , 7, 219-21.
(47) Harduf, Z.; Nir, T.; Juven, B. J. J. Chromatogr. 1 9 8 8 , 437, 379-86.
(48) Price, J. R.; Metz, P. A.; Veening, H. J. Chromatogr. 1 9 8 7 , 424, 795-9.
(49) Yun, Z.; Zhang, R. N. Biomed. Chromatogr. 1 9 8 7 , 2, 173-6.
(50) Sabri, M. I.; Soiefer, A. I.; Kisby, G. E.; Spencer, P. S. J. Neurosci. Methods
1 9 8 9 , 29, 27-31.
UV spectra were recorded on a Varian (Springvale, Australia)
Cary 3, IR spectra were run on a Perkin-Elmer Model 781
(51) Descombes, A. A.; Haerdi, W. Chromatographia 1 9 9 2 , 33, 83-6.
(52) Maeder, G.; Pelletier, M.; Haerdi, W. J. Chromatogr. 1 9 9 2 , 593, 9-14.
Analytical Chemistry, Vol. 68, No. 1, January 1, 1996 87

Figure 2. Chromatogram of the reaction mixture at 15 min. Peak
assignments are acetone (at 1.9 min), SMZ (3.7 min), FMOC-OH
(11.5 min), FMOC-SMZ (12.3 min), and unreacted FMOC (at 15.5
min).
Figure 3. Effect of pH on the reactions of FMOC: ], SMZ; [,
FMOC-SMZ; b, FMOC-OH. The relative areas were determined at
254 nm.
(Norwalk, CT), fluorescence measurements were performed on
a Shimadzu (Kyoto, Japan) RF-500 spectrofluorophotometer, and
NMR spectra were obtained from a Varian (Palo Alto, CA) NMR
Gemini 200 H/ C. Operating frequencies were 200.0 and 50.0
MHz.
P rocedures. As SMZ is not fluorescent, all of the studies of
rates and yields were necessarily performed using UV absorbency.
All reactions and analyses were performed at ambient tempera-
tures.
conditions were used as described in the previous section. After
20 min at each pH, the reaction mixture was analyzed at 254 nm.
Figure 3 shows the measured chromatographic areas for SMZ,
FMOC-SMZ, and FMOC-OH. It should be noted that the relative
areas of SMZ and FMOC-SMZ are not in the ratio of the molar
absorptivities stated in the last section of this paper due to
measurement at 254 nm rather than 264 nm. (Areas for FMOC
are not shown due to the overlap of the bands in the 13.4-15.8
min area of the chromatogram and the irreproducible integration
of the area.) The coincidental minimum in the SMZ residue and
maximum in the FMOC-SMZ adduct indicate about 8 to be the
optimum pH for reaction of FMOC with SMZ. A further study at
half pH unit intervals between pH 6 and pH 10 confirmed that
the sulfonamide amine group had maximum reactivity with FMOC
in the range pH 7.8-8.0, as do the amino acids.
Under the typical reaction conditions used aboves2 mL of
aqueous phases and 1 mL of acetone solutionsFMOC reagent
solutions above 15 mM led to precipitation of that reagent upon
mixing. Hence, the required stoichiometry of reaction was
established in a series of experiments in which the water content
of the reaction mixture was reduced. One milliliter of 4 mM SMZ
in pH 8 bicarbonate buffer was mixed with 4 mL of 1-15 mM
FMOC in acetone. Figure 4 shows the relative chromatographic
areas (at 254 nm) for FMOC-SMZ and FMOC-OH after 30 min
for concentrations of FMOC up to 8 mM and demonstrates the
need for a minimum ratio of FMOC/ SMZ as 7:1. Reaction can
be seen to be complete in 30 min.
RESULTS AND DISCUSSION
HP LC Analysis of Reaction Mixtures. Using the conditions
found to be optimal for the reaction of FMOC and amino acids as
determined by Einarsson’s study,39 and confirmed by Cunico et
al.,40 the reaction of FMOC and SMZ was first tested at pH 8.
One milliliter of 1 mg/ mL (3.59 mM) SMZ solution, 1 mL of pH
8 buffer, and 2 mL of 10 mM FMOC in acetone were mixed.
Preliminary studies indicated that the pH for RP-HPLC analyses
was not critical in the low end of the available range (2.5-4.5).
pH 3.5 was chosen. Isocratic analysis was impractical since the
weak elution conditions required for the separation of SMZ from
the large acetone solvent peak yielded unacceptably long retention
times for the other reagents and products, especially FMOC. The
analysis conditions found to be most satisfactory on the conven-
tional HPLC system were a medium concave gradient from 65%
A (0.05 M NaH2PO4, pH 3.5)/ 35% B (methanol) to 15% A/ 85% B
over 4 min, followed by isocratic elution with 85% methanol. The
flow rate was 1 mL min-1
. From comparison with SMZ and
A further constraint upon reaction also involves the FMOC
concentration. For the typical reaction conditions, initial concen-
trations of FMOC below 4 mM clearly led to the preeminence of
the hydrolysis reaction and incomplete derivatization of SMZ,
irrespective of its concentration. Hence, there are upper (solubil-
ity) and lower (rate of hydrolysis) limits to the acceptable FMOC
concentrations with respect to the water content of the reaction
mixture, as well as the need for a minimum ratio of FMOC/ SMZ
of 7:1.
FMOC blanks, prepared by substitution of the other reagent with
an equal volume of the appropriate solvent, unambiguous assign-
ments are possible for all reagents and products. Figure 2 shows
the chromatogram after 15 min of reaction. It can be seen that
baseline resolution was achieved for acetone, SMZ, the FMOC
hydrolysis product (FMOC-OH) formed during reaction, and
FMOC-SMZ. Unreacted FMOC at 15.5 min can be seen to have
undergone on-column hydrolysis in both the FMOC blank and
the reaction mixture, giving rise to the fronting structure compris-
ing the sharp peak at 13.6-13.7 min and the area up to 15.5 min.
Conditions for Derivatization. With the exception of the 1
mL of buffer, for which the pH was varied, the same reaction
Following isolation of the FMOC reaction products (see
below), a mass balance study versus time was undertaken. Using
the fast LC system, but with the same phases in reservoirs A and
88 Analytical Chemistry, Vol. 68, No. 1, January 1, 1996

Figure 4. Effect of the molar ratio of FMOC/SMZ on the extent of
reaction. Symbols and the absorption wavelength are as in Figure
3.
Figure 5. Variations in concentrations of the reagents and products
with time: ], SMZ; [, FMOC-SMZ; O, FMOC; and b, FMOC-OH.
B, the acetone solvent and SMZ were baseline resolved in 1.2
min at 30% methanol, and a jump to 70% methanol was effected
over 1.25 min. Chromatography similar to that shown in Figure
2 was obtained, and all peaks were eluted in 12 min. One milliliter
of ∼2 mM SMZ in bicarbonate buffer was reacted with 1 mL of
∼15 mM FMOC in acetone. Calibration curves for each of the
reagents and products were obtained concurrently with the
analyses. Good linear plots resulted: for SMZ (11 points, 0.02-
2.00 mM), R ) 0.9999; for FMOC-SMZ (6 points, 0.1-1.2 mM),
R ) 0.9984; for FMOC (14 points, 0.075-15.00 mM), R ) 0.9989;
and for FMOC-OH (11 points, 0.5-10 mM), R ) 0.9992. For the
residual FMOC concentration, only the chromatographic area due
to the unhydrolyzed reagent was used. The results are shown in
Figure 5. From the calibration curve and the SMZ blank, the
initial concentration of SMZ in the reaction mixture was deter-
mined as 0.91 mM. It is clear that the SMZ is quantitatively
converted to its FMOC derivative. As the reaction mixture was
initially 7.4 mM in FMOC and the final solution was 0.9(1) mM
in FMOC-SMZ, the expected final concentration of FMOC-OH
was 6.5 mM. Figure 5 indicates a final FMOC-OH concentration
around 6.7 mM. Given that there are several minor impurities in
the FMOC (that may be seen in the blank) and that the FMOC-
OH solid undergoes degradation, within experimental error, mass
balance does appear to be have been achieved.
Isolation and P urification of FMOC-SMZ and FMOC-OH.
FMOC-SMZ crystals were produced when 1 mL of pH 8 buffer
was mixed with 4 mL of 2 mM SMZ (in 50% acetone/ 50% water),
added to 5 mL of 15 mM FMOC in acetone, mixed well, and left
overnight. The solution was filtered with No. 4 filter paper, and
the crystals were washed with small amounts of acetone and then
distilled water (three times each) and left to dry in a vacuum
desiccator overnight.
FMOC-OH was produced by mixing 15 mM FMOC in acetone
with water in equal volumes and leaving overnight. The solution
was blown down to half its volume, filtered with No. 4 paper, and
then washed and dried as for FMOC-SMZ.
Figure 6. Infrared spectra of (a) SMZ, (b) FMOC-SMZ, and (c)
FMOC.
calculated) elemental compositions were C, 65.07% (64.79%); H,
4.95% (4.83%); N, 11.19% (11.19%); S, 6.20% (6.40%); and O (by
difference), 12.59% (12.79%).
IR Spectroscopy. Figure 6, parts a-c show the IR spectra in
KBr disks for SMZ, FMOC-SMZ, and FMOC, respectively.
Reference spectra in Nujol for SMZ and FMOC53 appear to be
identical other than for the dispersant.
Verification of Composition and Structure of FMOC-SMZ.
Microanalysis. Commercial microanalysis of the SMZ derivative
confirmed the composition to be C27H24O4N4S and the crystals to
have a clear-cut melting point of 258 °C. The microanalytical (and
(53) Pouchert, C. J. The Aldrich Library of FT-IR Spectra; Aldrich Chemical Co.:
Milwaukee, WI, 1985.
Analytical Chemistry, Vol. 68, No. 1, January 1, 1996 89

Reference to Colthup, Daly, and Wiberley54 permits a full
interpretation of the spectral changes observed. First, there are
clear changes to all absorbances associated with SMZ vibrational
modes near the reaction site. Above 3100 cm-1, SMZ dislays the
typical structure of three sharp, medium-intensity peaks associated
with primary aromatic amines. Absorption at 3435 cm-1 is due
to the N-H asymmetric stretch. The other two absorbances at
3335 and 3235 cm-1 are a result of Fermi resonance between the
N-H symmetric stretch (expected at 3400-3200 cm-1) and the
weak overtone of the NH2 deformation (at 1635 cm-1) that would
be expected around 3270 cm-1. Because these three peaks are
all dependent on the primary amine stretching fundamentals, all
are missing from the FMOC-SMZ FT-IR spectrum. They are
replaced by a broad, medium-intensity band at 3280 cm-1 (which
is compatible with the absorbance expected at 3400-3070 cm-1
for monosubstituted amides) and a weak, medium-intensity broad
band centered on 3000 cm-1. Again, this structure is probably
due to Fermi resonance between an overtone (of the 1550 cm-1
band in this case) and a stretching fundamental (3400-3070
cm-1).
the C-O bond, and shift it to lower frequencies (relative to an
adjacent C), while the adjacent N in the derivative is known to
have a net electron donation effect, thus strengthening the C-O
bond and shifting it to a higher absorbance frequency.
A band in the 694-689 cm-1 range is stated to be the C-Cl
stretch in chloroformates. Examples of other chloroformates
confirm the presence of a medium-intensity peak just below 700
cm-1
. However, allowing for dilution of the FMOC in the
derivative, there appears to be little loss of intensity at this
frequency. The matter is further complicated by the presence of
a medium-intensity absorbance in about the same place in the
spectrum of fluorene.53
The third and final source of IR absorption evidence for the
reaction of SMZ as shown in Figure 1 is provided by a comparison
with spectra of molecules with the same structure as FMOC-
SMZ in the region of reaction. Monosubstituted amides in general
provide a reference pattern; phenyl-substituted amides provide a
more specific reference, and, as indicated above, carbamates
provide the model compounds.
The weak SMZ band at 1545 cm-1 (of the sulfonamide N-H
out-of-plane bend) is swamped in FMOC-SMZ (1535 cm-1) due
to the superimposition of the complex CNH vibration that is very
characteristic for monosubstituted amides (near 1550 cm-1). The
next lowest absorbance frequency of significance is centered about
1045 cm-1 in FMOC-SMZ. This medium-strong absorbance is
absent in both reagents and would therefore be expected to be
closely associated with the newly formed C-N bond, especially
as this falls within the normal range of absorbance for C-N
stretching. Inspection of carbamate R1NHCOOR2 IR spectra54
rules out this possibility, and the authors assign the absorbance
to the O-R2 bond. The correctness of this appears to be verified
by the spectrum of the compound PhNHCONH2, in which there
For a primary aromatic amine, the NH2 deformation is expected
at 1650-1590 cm-1. Hence, the absence of the 1635 cm-1 SMZ
peak in the FMOC-SMZ spectrum confirms that NH2 is the
reacted SMZ functional group. Primary aromatic amines also
absorb strongly at 1330-1260 cm-1 due to stretching of the phenyl
(Ph) carbon-nitrogen bond, in this case at 1300 cm-1 in the SMZ
spectrum. In FMOC-SMZ, the strong Ph-N stretching absor-
bance would be expected to move to higher wavenumbers.55 It
is observed at 1310 cm-1 and is again consistent with the
derivatization.
At 675 cm-1 in SMZ only, the strong absorption is due to the
-NH2 wag or rock (but is not broad, as is often shown for the
spectrum of the liquid). The removal of this peak again indicates
derivatization of the primary amine in SMZ. The comparable out-
of-plane -NH wag for N-substituted amides absorbs broadly near
are no significant absorbances below 1160 cm-1
. For the five
carbamate spectra given, the O-C stretching frequency varies
from 1120 to 1030 cm-1. Of greatest interest is the spectrum of
ethyl N-phenylcarbamate, PhNHCOOC2H5. This compound is the
closest analogue to FMOC-SMZ and has the O-C stretch
assigned at 1070 cm-1. It also has the CO-O absorption assigned
to a very strong band at about 1240 cm-1. This is equivalent to
the FMOC-SMZ absorbance at 1200 cm-1. It appears that the
two C-O absorbances in FMOC-SMZ (CO-O, 1200 cm-1 and
O-CH2, 1045 cm-1) are merged in the FMOC reagent. This
would also be consistent with the very high intensity of the FMOC
1140 cm-1 absorbance.
700 cm-1 54 This appears to be centered on 730 cm-1 beneath
.
various FMOC absorbances.
The second series of spectral changes that support the
proposed reaction are those observed between FMOC and
FMOC-SMZ. In chloroformates, carbonyl absorbances are
expected at 1860-1760 cm-1; in FMOC, this (split) peak is at
1775-1750 cm-1
. The relative effects of chloro and amino
substitution upon (CdO) absorbances in ketones are well known.56
Relative to an unsubstituted ketone, chloro substitution causes a
shift to higher frequencies, while amino substitution lengthens
and weakens the bond, leading to lower frequencies. Thus, the
CdO absorbance frequency should be lower in the derivative. The
FMOC-SMZ CdO absorbance occurs at 1730 cm-1 and is
consistent with the observed range for phenyl N-substituted
carbamates (1730-1700 cm-1).
One further feature of the ethyl N-phenylcarbamate spectrum
is the N-H stretch. It is a broad, medium-intensity band centered
just above 3300 cm-1 and is very similar to the FMOC-SMZ band
centered just below 3300 cm-1
.
NMR Spectroscopy. The NMR spectra of SMZ, FMOC-SMZ,
and FMOC were obtained in dimethyl sulfoxide and are shown
in Figure 7, parts a-c, respectively. For SMZ, the NMR spectrum
is in excellent agreement with the literature.57 For FMOC, a
reference spectrum was not located, but that of fluorene57
confirmed the majority of absorbances. (The broad absorbance
centered on 8.25 ppm is assumed to be due to the readily formed
hydrolysis product.)
Chloroformates have a C-O stretch at 1172-1134 cm-1. It is
observed at 1140 cm-1. In the SMZ derivative, this very strong
peak can be seen to have shifted to a higher frequency, 1200 cm-1
.
This is again consistent with the derivatization. The adjacent (Cl)
group in the FMOC reagent would withdraw electrons, weaken
(54) Colthup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and
Raman Spectroscopy, 3rd ed.; Academic: London, 1990.
(55) Pecsok, R. L.; Shields, L. D.; Cairns, T.; Mcwilliam, I. G. Modern methods of
chemical Analysis, 2nd ed.; John Wiley: New York, 1976; p 70.
(57) Pouchert, C. J.; Campbell, J. R. The Aldrich Library of NMR Spectra; Aldrich
(56) Kemp, W. Organic spectroscopy, 2nd ed.; Macmillan: London, 1987.
Chemical Co.: Milwaukee, WI, 1974.
90 Analytical Chemistry, Vol. 68, No. 1, January 1, 1996

Figure 7. 1H NMR spectra of (a) SMZ, (b) FMCO-SMZ, and (c)
FMOC.
In the SMZ spectrum, the singlet equivalent in area to two
protons at δ ) 5.94 is due to the two amine protons.58 This
absorbance is missing in the spectrum of the derivative. The
remaining proton at this site (-OCONH-) is downshifted by the
deshielding from the OCO group and from the areas is seen to
be located within the aromatic group at δ ) 7.2-7.4. This is also
consistent with the tabulated range of chemical shifts for phenyl-
substituted amides58 at 7.5-8.5 ppm. The other revealing assign-
ment is for the SMZ two-proton doublet at δ ) 6.5-6.6, which is
also absent in the spectrum of the FMOC adduct. Taking benzene
as the standard substance with δ ) 7.27, the protons ortho to the
amine group may be assigned via substituent effects:58 δ ) 7.27-
0.8 [o-NH2] + 0.15 [m-SO2] ) 6.62 ppm. Replacement of the
amine substituent by a monosubstituted amide (as an approxima-
tion) leads to δ ) 7.27 + 0.4 [o-NCOR] + 0.15 [m-SO2] ) 7.82
ppM, and comparison of the FMOC-SMZ spectrum with that of
FMOC shows a 2-fold increase in area of the FMOC two-proton
doublet at δ ) 7.2-7.4.
In the FMOC spectrum, for the 9-fluorenyl proton, δ ) 1.54
[CH] + 1.3 [Ph] × 2 ) 4.14 ppm, and a triplet equivalent to one
proton is located between 4.0 and 4.1 ppM. For the adjacent
methylene protons, δ ) 1.25 [CH2] + 2.7 [OCO] ) 3.95 ppM,
and the required doublet equivalent to two protons is found
between 3.7 and 3.8 ppM. As these protons are adjacent to the
reaction site, their chemical shifts would be expected to change.
From the peak areas and splittings, it is clear that the signals
generated around 4.5 ppM in the FMOC-SMZ spectrum are due
to the CH2 protons, and the CH proton absorbs at 4.4 ppM. This
inversion of order of the chemical shifts is not surprising, since
the downfield shift caused by SMZ is weakened by distance.
In summary, the structure of FMOC-SMZ analyzed from the
NMR spectrum is coincident with the expected one.
Figure 8. Fluorescent response of FMOC-SMZ in the 100-1000
ppt range.
FMOC-OH. In 70% pH 3.5 phosphate buffer/ 30% methanol, the
values were 264 and 20 250 for SMZ. In addition, ꢀ was measured
at exactly 264 nm in the 70% methanol phase for SMZ (20 180
mol-1 L cm) and FMOC-SMZ (53 760 mol-1 L cm). These data
show that for UV absorbance detection in the vicinity of the peak
maxima, the sensitivity for SMZ may be increased by a factor of
2.7 by derivatization with FMOC and that the absorptivity of the
derivative (53 760) is enhanced relative to those of the reagents
(20 180 + 19 810 ) 39 990). As the absorbance centers are
independent in the derivative and the fluorene moiety is distant
from the reaction site, this enhancement must arise from an
increase in absorbency of the sulfonamide phenyl system.
The fluorescence spectra of FMOC and its derivatives are well
known.40 Maximum absorbance in excitation occurs at 264 nm,
and in emission a maximum is observed around 307 nm. FMOC-
SMZ was diluted in the 70% methanol buffer. Figure 8 shows the
fluorescence response in the lowest concentration range exam-
ined. A linear response is evident, R ) 0.924. In the ppb range
(10 points, 10-100 ppb, R ) 0.9976), less scatter was observed
and linearity maintained. However, when the concentration is in
the 1-10 ppm range, gentle curvature begins to appear in the
fluorescence response. Above this, the response rapidly maxi-
mizes (at 20 ppm) and then “exponentially” decreases up to 100
ppm.
UV Absorption and Fluorescence Spectroscopy. UV ab-
sorbance spectra were determined for both of the reagents and
both of the products at 0.04 mM in the mobile phases in which
they were monitored during the mass balance study. In 30% pH
3.5 phosphate buffer/ 70% methanol, the wavelengths of maximum
Without detection limit studies, exact sensitivities cannot be
determined. However, an estimate of the increase in sensitivity
is obtained from spectrofluorometric detection of the FMOC-
SMZ derivative and spectrophotometric UV absorbance of SMZ
when standard 1 cm path length/ 3 cm3 cells were used. Taking
the 0.81 absorbance of 0.04 mM SMZ, assuming linearity and the
minimum detectable absorbance to be 0.005, the minimum
detectable concentration of SMZ by UV absorbance would be 60
ppb. From Figure 8, it is clear that an improvement of 3-4 orders
absorbance (λmax)/ nm and the molar absorptivities (ꢀ)/ mol-1
L
cm at those wavelengths were determined as 264 and 19 810 for
FMOC, 262 and 54 250 for FMOC-SMZ, and, 265 and 19 110 for
(58) Kemp, W. NMR in Chemistry. A Multinuclear Introduction; Macmillan:
London, 1986.
Analytical Chemistry, Vol. 68, No. 1, January 1, 1996 91

of magnitude is possible by fluorence detection of the FMOC
derivative.
columns from the Varian Instrument Division, Walnut Creek. We
thank Ross Mair for running the NMR spectra.
Received for review May 31, 1995. Accepted September
13, 1995.X
ACKNOWLEDGMENT
G.S.L. and Z.Y.Z. acknowledge the receipt of Overseas Post-
graduate Research Awards from the Australian Government. We
gratefully acknowledge the donation of the fast LC system and
AC9505280
X Abstract published in Advance ACS Abstracts, November 15, 1995.
92 Analytical Chemistry, Vol. 68, No. 1, January 1, 1996