REVIEW OF ASCORBIC ACID METHODOLOGY

Analytical Methods for Determining Ascorbic Acid in Biological Samples, Food Products, and
Pharmaceuticals
LAWRENCE A. PACHLA and DONALD L. REYNOLDS
Warner-Lamberti'Parke-Davis, Pharmaceutical Research, PharmacokineticslDrug Metabolism 2800 Plymouth
Rd, Ann Arbor, MI 48105
PETER T.KISSINGER
Purdue University, Chemistry Department, West Lafayette, IN 47907

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

Over the last decade, numerous publications have appeared describing/ \
analysis (3, 4). A symposium on ascorbic acid, its chemistry,
analyses for ascorbic acid in food products, pharmaceuticals, and metabolism, and uses, was sponsored by the Division of
biological samples. This review focuses on the chemistry associatedCarbohydrate Chemistry at the Second Chemical Congress
with many of these procedures. The papers discussed have historicalin 1980. At this symposium, Sauberlich and colleagues pre-
importance, are important to understanding the method, or have sig-sented a review that discussed the advantages of liquid chro-
nificantly advanced ascorbic acid analysis. The review has 4 major,
sections: spectroscopic, electrochemical, enzymatic, and chromato- matographic (LC) methods over spectroscopic methods (5).
graphic methods of analysis. Methods in Enzyrnology has also devoted a section in a recent
volume to ascorbic acid (6).
Ascorbic acid is an important vitamin having a chemical Recently, Cooke and Moxon have reviewed spectroscopic
structure that justifies its classification as a carbohydrate. Its and chromatographic methods of analysis for this vitamin in
physical/chemical properties, compiled by Kutsky (1), are as food products (7). That review does not include electrochem-
follows: aqueous solubility = 0.3 g/mL, melting point = 190- ^ ical or enzymatic methods of analysis.
192°C, redox potential = E0 = 0.166 V at pH 4, pKa = 4.17, This review surveys the analysis for ascorbic acid in phar-
pKa2 = 11.57, and absorption maxima = 245 nm (acid) and maceuticals, food products, and biological samples, and focuses
265 nm (neutral). Ascorbic acid is rapidly oxidized to dehy- on the chemistry associated with many of these procedures.
droascorbic acid (DHAA) by oxygen and metal ions in alka- A review of stability or sample extraction has not been included,
line and acidic media. Dehydroascorbic acid can be further because this topic is discussed in cited papers or the above
oxidized to diketogulonic acid. The structures of these 3 reviews. Only selected papers of historical importance have
compounds are shown below: been cited, those which are important to the chemical under-

C00H
*
t

KK-H
H-C-OH
CH2OH H,i~0H

Ascorbic Acid Dehydroascorbic Acid Diketogulonic Acid

Over the last decade, numerous publications have appeared standing of the method, or important in advancing the selec-
describing analyses for ascorbic acid in food products, phar- tive, sensitive analysis for ascorbic acid. The review has been
maceuticals, and biological samples. This resurgence in new divided into 4 major sections: spectroscopic, electrochemi-
analytical procedures results from the importance of the vita- cal, enzymatic, and chromatographic methods of analysis.
min in nutritional, clinical, pharmacological,.and industrial"
studies. A monograph containing selected procedures for Spectroscopic Methods
pharmaceutical preparations has appeared (2). In addition, a The classical chemistry associated with spectroscopic
review on the determination of this vitamin in pharmaceuti- methods can be divided into 2 general categories: (a) those
cals has been included in biannual updates on pharmaceutical using a redox indicator in its oxidized form, and (b) those
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68,. NO. 1, 1985)

Lawrence A. Pachia is a research Donald L. Reynolds is a scientist Peter T. Kissinger is a professor

associate in the in the Pharmacokinetics/Drug of chemistry at Purdue

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

Pharmacokinetics/ Drug Metabolism Department of University and president of
Metabolism Department at Warner-Lambert/Parke-Davis. He Bioanalytical Systems, Inc.
Warner-Lambert/Parke-Davis received his B.S. degree in (BAS), which he founded in
and is a lecturer in the Pharmacy from Ferris State 1974. He received a B.S. in
Chemistry Department at College in 1976, and his Ph.D. Chemistry (1966) from Union
Lawrence Institute of degree in Pharmaceutical College, Schenectady, and a
Technology. He received his B.S. Chemistry from the University of Ph.D. in Analytical Chemistry
degree in Chemistry with a Kansas in 1983. His research (1970) from the University of
minor in Physics in 1973 and his interests include pharmaceutical North Carolina at Chapel Hill,
Ph.D. in Analytical Chemistry analyses, bioanalytical working with the late Professor
with a minor in Biochemistry chemistry, and Charles N. Reilley. Before
from Purdue University in 1978: pharmacokinetics.. He has joining the faculty at Purdue in
From1978 to 1981 he was publications concerning the 1975, Kissinger was a research
employed in the analytical and theory of liquid chromatography associate at the University of
bioavailability/pharmacokinetics and its application to Kansas (1970-72) and an
groups at McNeil pharmaceutical analysis. He is a assistant professor at Michigan
Pharmaceutical. His research member of the American State University (1972-75). His
interests are the automation of Pharmaceutical Association and research has involved the study
analytical procedures, the Academy of Pharmaceutical of organic electrode reaction
bioanalytical and Sciences. mechanisms,
electroanalytical chemistry, spectroelecirochemistry,
application of chromatographic hydrodynamic electroanalytical
procedures to drug analysis, techniques, modern liquid
pre-column and post-column chromatographic techniques,
derivatization techniques, and and electrochemical
the areas of pharmacokinetics methodology for the
and bioavailability. He has neurosciences. Most recently,
published papers on the Kissinger research group
pharmaceutical, clinical, and has been investigating the
food analysis. He is a member of metabolism of aromatic
the American Pharmaceutical xenobiotics with the goal of
Association, American Chemical better understanding the
Society (ACS), and ACS molecular basis of toxicity.
Analytical Chemistry and Kissinger and colleagues have
Chromatogrphy Divisions. published over 115 research
articles and presented over 200
invited lectures in 15 countries.
He is a founding member of the
Society for Electroanalytical
Chemistry (SEAC), and is active
in the American Chemical
Society and the American
Association for Clinical
Chemistry.
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985) 3

involving chromogen formation by derivatization. The redox phenol, and therefore interfere. Attempts to minimize these
indicators that will be discussed include 2,6-dichloroindo-. positive biases have included chromatographic isolation of
phenol, metal ions, and other miscellaneous reagents. The ascorbic acid (22) and heavy metal ion removal via mercap-
derivatization of ascorbic acid includes reaction with 2,4- tide salt precipitation before the reaction (23). Another way
dinitrophenylhydrazine, diazotization, and quinoxaline for- to improve the specificity involves following the kinetics of
mation. the reaction between the vitamin and DCIP. Hiromi and
coworkers used the linear relationship between the apparent
Redox Reactions first order rate constant for reduction of the indophenol to
2,6-Dichloroindophenol.—The standard redox reagent used determine ascorbic acid concentration (24, 25). The reaction
for ascorbic acid analysis in a variety of sample types is 2,6- of excess ascorbic acid with DCIP was linear between 0.5
dichloroindophenol (DCIP). Solutions of the reagent are blue |xM and 50 (JLM and was suitable for analysis of orange juice.
at neutral pH and pink in acid. The stoichiometry of the redox Cysteine and glutathione did not interfere, and interference
reaction was first proposed by Tillmans (8). The reaction is by triose reductone was minimized. Recently, these investi-
rapid and is first order with respect to each reactant. The gators improved the lower detection limit to 2 (xM by reacting
overall second order rate constant was pH-dependent with ascorbic acid in the presence of excess reagent and using a
an optimum value of 56.5 x 103 L/mole/s at low pH (9, 10). stop-flow kinetic procedure (26). This method was applied to

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

CI

->-DHAA

CI

The first application of 2,6-dichloroindophenol to ascorbic a variety of vegetables and fruits and requires only 1-2 s/
acid analysis was reported in 1932 by Tillmans et al. (11). sample. The suitability of this type of methodology depends
After numerous modifications, the method still consists of on advances in stop-flow methods (27) and a need to analyze
monitoring DCIP absorbance at 518 nm before and after sam- many samples. This methodology should find acceptance in
ple addition. Hoffman et al. have automated the DCIP pro- clinical laboratories with centrifugal analyzers.
cedure and used it to determine the concentration of ascorbic Metal Ions.—Another class of colorimetric redox reactions
acid in orange and grapefruit juices (12). Their method was involves the reduction of metal ions to produce a stable col-
unsuitable for the assay of lemons because of the high acidity ored solution. For example, ascorbic acid may be determined
of the sample. Sample turbidity was minimized by dialyzing in pharmaceutical preparations by using potassium ferricyan-
the ascorbic acid from the sample directly into the flowing ide (28) or ferricinium trichloroacetate (29). Pelizzeti. et al.
DCIP solution. Another automated version introduced by (30, 31) have reported the kinetics and mechanisms between
Egberg and coworkers was applied to the assay of fortified ascorbic acid and many potentially useful metal-ion complex
grains and beverages (13). This method required a metha- oxidants.
nolic-metaphosphoric acid extraction step to reduce turbid- The more common metal-ion redox methods involve the
ity from colloidal suspensions. The detection limit of the reduction of iron(III) to iron(II) by the vitamin. An intensely
method was 6 fxg/mL and response was linear up to 200 |xg/ colored iron(II) complex is formed after addition of a chelat-
mL. ing agent. Absorbance is monitored at the absorption maxi-
There has been considerable interest in determining ascor- mum of the complex and is directly proportional to the ascor-
bate concentrations for nutritional and biomedical surveys. bic acid concentration. The most common iron(II) chelating
Because of the large number of samples in these studies, agents are a,a'-dipyridine, 2,4,6-tripyridyl-S-triazine (TPZ),
several variations of the basic methodology have been devel- and ferrozine.
oped. Methods have been introduced for cataractous lens Zannoni and coworkers have published a micromethod for
(14), ovarian (15), and lingual tissues-(16); infant milk (17); the determination of ascorbic acid in plasma and tissue sam-
and serum, whole blood, and saliva (18-20). One automated ples (32). This procedure detects as little as 0.1 fxg ascorbic
version, rigorously evaluated by independent investigators, acid and requires 0.01 mL sample; the complex is stable for
has been chosen as a "selected method" for clinical labora- at least 2 li\ Okamura has reported a procedure for the
tories (21). This method produces a linear response from 2.0 simultaneous determination of ascorbic and dehydroascorbic
to 20 jxg ascorbic acid/mL serum, with a maximal sampling acidsin plasma (33,34). In both procedures, dehydroascorbic
rate of 60/h. Interferences from sulfhydryls, sulfites, and thio- acid is reduced to ascorbic acid with dithiothreitol. The ascor-
sulfates and from catalytic oxidation of ascorbate by iron(III) bate and dehydroascorbate: concentrations are then deter-
and copper(II) were minimized by using a pH 3.5 solution of mined by difference. Precipitation of iron from urine samples
metaphosphoric acid. '; ; and treatment with activated charcoal to remove interfering
The limitations encountered with DCIP methods result from substances were necessary. Excellent agreement was found
poor specificity and dye instability. Many molecules (e. g., between this procedure and a dinitrophenylhydrazine method.
phenols, sulfhydryls, and triose reductones) and ions (e.g., Lloyd et al. have used 2,4,6-tripyridyl-5'-triazine to develop
ferrous, cuprous, or sulfite) can reduce 2,6-dichloroindo- a procedure for analysis for ascorbic acid in platelets (35).
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)

CH3O DHAA +' OCH.-

or
€L OCH3

CH3O

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

The platelets are extracted and the supernate is chromato- by ascorbic acid (43). The molar absorptivity of the product
graphed on cellulose TLC plates. Ascorbic acid was eluted is 1738 at 510 nm and Beer's law is followed for ascorbate
and reacted with iron(III) in the presence of the complexo- concentrations between 10 and 80 [ig/mL. This reagent has
metric reagent. The precision of this method was 2.5%. This been modified and applied to the quantitation of the vitamin
reagent was also suitable for the analysis of plasma samples in citrus fruits (44). The reaction has been characterized (45)
(36). Acidification of plasma followed by reaction with iron(III) and is complete within 6 min at 25°C for amounts of ascorbic
and treatment with chelating reagent are the only necessary acid as high as 0.8 mg. The reaction product has been syn
treatment steps. This procedure was applicable to clinical thesized and characterized by infrared, nuclear magnetic res
samples. onance, and mass spectrometric techniques. The reduced
Another excellent iron(II) complexing agent in widespread dimethoxyquinone was present as 2 isomeric products.
use is ferrozine (3-(2-pyridyl-5,6-bis(phenylsulfonic acid)-l,2,4- Kinetic analytical methods have appeared that use bromine
triazine)). The advantages of ferrozine over other Fe(II) generated from the reaction between bromate and bromide
reagents include water solubility, stability, and a higher molar in acidic solution (46, 47) or oxidation of ascorbic acid by
absorptivity (27 900) when compared with phenanthroline iodate (48). White and Fitzgerald used methylene blue to
(11 000) or a,a'-bipyridyl (8650) complexes. The kinetics of analyze ascorbic acid tablets and fruit juices (49). The reduc-
the reaction between iron(II) and ferrozine, 1-10-phenanthro- tion of perinaphthindan-2,3,4-trione at pH 3 with subsequent
line, and a,a'-bipyridine have been compared (37). Jaselskis monitoring of the reduced chromogen at 460 nm was the basis
and Nelapaty have characterized the optimal reaction con- of another method used for citrus fruits (50). The ease of
ditions and have developed a method suitable for the quan- determining the titer of standard ascorbate solution has been
titation of ascorbic acid in citrus fruits (38). A pH 3-6 solution facilitated by using o-toluidine (51), and phosphotungstic acid
was optimal for color formation, and as little as 0.2 (xg ascor- has been used as the basis of a serum ascorbate method (52).
bic acid/mL could be quantitated. Butts and Mulvihill have Two phosphotungstic acid methods have been published for
used this chromogenic reagent to develop a procedure for the determination of ascorbic acid in pharmaceuticals, plasma
determining the vitamin in serum or urine (39). Their proce- (53), and vegetables (54). Folin phenol reagent has been used
dure was based on rapid formation of the ferrozine complex, for determining ascorbic acid in biological samples (55).
using the unique ability of a centrifugal analyzer to simulta-
neously measure standards and samples (and a blank if nec- Derivatization Reactions
essary) after very short reaction times (>15 s). By carefully
Dinitrophenyl Hydrazine.—Another classical colorimetric
optimizing the pH, reagent composition, and reaction time,
method involves measuring the absorbance produced when
they were able to successfully assay for ascorbic acid in the
dinitrophenylhydrazine (DNPH) couples with the oxidized
presence of uric acid, sulfhydryls, and other reductants.
form of ascorbic acid. In this procedure (commonly termed
Recently, this procedure has been modified for manual deter-
the Roe and Kuether method) (56), ascorbic acid is first con
mination of the vitamin in tissue samples (40).
verted to dehydroascorbic acid (DHAA) via a suitable oxi
Other reported metal-chelometric procedures include the dizing agent. Dinitrophenylhydrazine is added and, after a
formation of an iron(II)-bathophenanthroline complex to brief incubation time, color is produced after acidification
determine ascorbic and dehydroascorbic acids in rat tissue with concentrated sulfuric acid. The absorbance is measured
samples (41), and formation of a copper(I)-2,2 bisquinoline at 520 nm and is proportional to the original ascorbic acid
complex which is the basis of a method for pharmaceutical and DHAA concentrations.
preparations (42). In general, the redox chelometric methods The specificity of the method is attributed to the following
have the same disadvantages as DCIP methods. Because the factors: (1) color is produced more easily with 2,4-dinitro-
vitamin is measured indirectly, the presence of other reducing phenylhydrazine derivatives of 5- and 6-carbon, sugar-like
agents will positively bias the results. •..-.," compounds; (2) the rate of coupling is much faster with dehy-
Miscellaneous Reagents.—Several other redox reagents droascorbic acid than with other carbohydrates; and (3) mea-
have been used for determining ascorbic acid in biological surable chromogen formation from non-ascorbic acid sub-
fluids, pharmaceuticals, and food products. One method for stances is minimized by carrying out the reaction at low
pharmaceutical preparations is based on the production of an temperatures. Zloch and coworkers (57) have evaluated the
intense red-violet color when dimethoxyquinone is reduced Roe and Kuether method using thin layer chromatography!
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)
5

mild
°2»v
AA > B H M + 2 H
% S °4
oKidatioa 2^^\ ^*>2

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

2
2 CHOH
I
CH2OH

and 14C-labeled ascorbic acid and found that only 30-53% of column has also been used for quantitation of ascorbic, dehy-
the bis-2,4,-dinitrophenylhydrazone derivative was pro- droascorbic, and 2-ketogulonic acids (65). A 50 mM H3P04
duced. Optimal derivatization conditions yielding maximal mobile phase was optimal.
derivative formation were 3-4 h at 37°C, or 1.5 h at 60°C.
The DPNH method has been automated by Aeschbacher
This approach has been used to monitor urinary ascorbate
levels (58). and Brown and applied to analysis of tissue samples (66). A
manual method has also been reported for analysis of platelets
A method has been reported, using the oxidant potassium and leukocytes for ascorbic acid (67). Serious disadvantages
bromate, that can quantitate ascorbic acid and its important of the manual methods include lengthy derivatization steps
metabolite,.ascorbic acid-2-sulfate, in urine and tissues (59). and the fact that DHAA is also measured as ascorbic acid.
Free ascorbic acid is quantitated after 1 h incubation at 37°C, Pelletier has reported a combined manual method incorpor-
although its metabolite is measured after incubation for 3 h ating dichloroindophenol and dinitrophenylhydrazine to dif-
at 60°C. The difference in absorbance at both temperatures ferentially measure ascorbic acid and DHAA levels in bio-
corresponds to the initial amount of ascorbic acid-2-sulfate. logical samples (68). An automated version of this method-
Many investigators have integrated selective chromato- ology has been adapted for determining ascorbic acid in serum
graphic techniques to increase specificity. For example, Bel- (69) and pharmaceutical products (70). An in-depth investi-
jaars et al. combined thin layer chromatography with the 2,4- gation of the automated experimental parameters has been
DNPH method (60). They applied densitometry detection for undertaken and applied to foodstuffs (71). The parameters
measuring 0.08-1.00 (xg ascorbic acid in buttermilk. Shmidt investigated included maximization of the conversion of DHAA
and Holfelder used thin layer chromatography to isolate the to ascorbic acid for blanks, minimization of osazone forma-
2,4-dinitrophenylhydrazone derivative from interfering com- tion by sugars, and optimization of color development. The
ponents and were able to quantitate the ascorbic acid content procedure was applied to vegetables, fruit juices, and other
in black current juice (61). Micro and macro methods using food products. The method was more specific when com-
this approach were developed to determine ascorbic acid pared with a manual DPNH method. Behrens and Madere
levels in bodyfluidsand tissues of insects (62). (72) have made further improvements in Pelletier and Bras-
Toothill and coworkers incorporated column chromatog- sard's method (69). Refinements in the reaction conditions
raphy and compared their results with a 2,6-dichloroindo- and choice of reagents allowed determination of ascorbic acid
phenol method for ascorbic acid analysis in evaporated and in small volumes of rat plasma and tissue extracts. The new
fortified sterilized milk (63); they reported the DPNH method procedure requires only 0.15 mL sample and could measure
to be more accurate. The presence of sulfhydryl groups and as little as 1.2 jxg/mL. Use of the method yielded standard
reductones present in evaporated and sterilized milk inter- deviations of 0.5-1.5% for ascorbate concentrations between
fered in the 2,6-dichloroindophenol method, but not in the 1.6 and 6.8 |xg/mL.
TLC/DPNH method. Dried feeds have been assayed for
ascorbate content with combined column chromatography Wahba and coworkers developed a simplified colorimetric
and the DPNH method (64). This approach has a throughput assay to determine ascorbic acid in pharmaceutical prepara-
of 50-60 samples/day and a detection limit of 0.1 u,g ascorbic tions (73). This method consists of reacting ascorbic acid with
acid/g. Sample cleanup on a Dowex 1-X2 anion-exchange phenylhydrazinium chloride in acidic solution at 50°C, with
subsequent measurement at 395 nm. The assay obeys.Beer's
 6 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)

law from 25 to 100 u,g, and interference by other vitamins The manual OPDA method has been automated on Tech-
and reducing substances is minimized. nicon Autoanalyzer by simply replacing the Norit oxidation
step with redox reagents (84-86), but a major drawback of
Diazotization.—Another class of derivatization methods using redox reagents in a totally automated system is that
involves the reaction of ascorbic acid with diazotized 4-meth- naturally occurring fluorescing plant components may inter-
oxy-2-nitroaniline. First studied by Schmall and coworkers fere because they are no longer adsorbed onto the Norit. This
(74, 75), the reaction mechanism is a hybrid redox-derivati- problem can be minimized by using a semiautomated version
zation and is illustrated below. (87).

O O n n „~ HO
ft III I |°2
-H«C-C-0

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

HO-cfi^O , U>
H,c 0
CH2OH

Derivatization is rapid, and the absorbance of the alkaline The automated method of Roy et al. (85) and semiauto-
blue solution (570 nm) is proportional to the original ascorbic mated method of Egberg et al. (87) have been compared (88)
acid concentration. Other vitamins and dehydroascorbic acid with 2 manual AOAC titrimetric and fluorometric methods i
do not react with the diazotization reagent. This methodology (89,90). In this study, 40 products (cereals, fruits, vegetables, j
is, therefore, appropriate for pharmaceutical stability studies. baby foods, meats, frozen dinners, juices, nutritional health I
The basic manual procedures have been modified and bars, and pet foods) were assayed in duplicate on 2 days and I
adapted to a Technicon Autoanalyzer for automated serum included a recovery study. Egberg's method was found to
and urinary ascorbate analysis (76). The automated method have a correlation factor of 0.999 with the official methods
obeys Beer's law over the concentration range of 1-7 mg/100 and an average recovery of 97.8%; while the Roy method had
mL, and specificity is enhanced for biological samples by a correlation factor of 0.979 and an average recovery of 99.3%.
using blank correction. Diazotized 4-methoxy-2-nitroaniline On the basis of these results, it has been suggested that
has also been used in an ovarian ascorbic depletion assay (77) Egberg's method is the method of choice for most of the I
and more recently in an automated pharmaceutical formula- products tested. A recent study has determined thewithin-
tion assay (78). Chromatography on a cellulose column has and between-laboratory reproducibilities of Egberg's semiau-
been used for sample cleanup for selective determination of tomated method (91). Fifteen samples of 12 different products
ascorbic acid in foodstuffs (79). Other variations of the method (cereals, fruit juices, and infant formula) were assayed by 5
use p-nitroaniline (80) and/or diazotized/?-aminobenzoic acid different laboratories. The within-laboratory relative stan-
as the redox-derivatization reagent (81). dard deviations for 3 different blind duplicate samples were
Quinoxaline Formation.—Alternative derivatization reac- 11.9,0.93, and 3.2%. The average relative standard deviation
tions include the formation of condensation products derived between laboratories was 4.9% (range 1.5-12.6%). On the]
between dehydroascorbic acid and substituted o-phenylene- basis of these results and the historical performance of the ]
diamines (OPDA): method (88, 91), the semiautomated version was adopted asj

AA DHAA
:r; CHOHCH2OH

The most commonly used method is the manual Deiitsch an official method at the 97th annual international meeting of
and Weeks assay (82). Norit (carbon) oxidizes ascorbic acid AOAC.
to DHAA which then reacts with ophenylenediamine to yield Other variations of the basic o-phenylenediamine approach
a fluorescent quinoxaline derivative. This method is more have used substituted analogs of OPDA in the reaction. For
specific, faster, and less restrictive to sample type when com- example, Szepesi used 4,5-dimethyl-l,2-phenylenediamine|
pared with the dichloroindophenol and DPNH assays (83). (92). This new derivatization reagent allowed the ascorbic I
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO.. 1,-1985) 7

acid content of pharmaceutical preparations to be determined samples/h), it could be readily automated using recent advances
by ultraviolet spectroscopy, and a 4-fold decrease in deriva- in automated chromatographic sample processors.
tization time was reported. Recently, Strohl and Curran (103) used flow injection anal-
Another spectrophotometric method that incorporates an ysis with a reticulated vitreous carbon flow-through electrode
analog of OPDA, 4-nitro-o-phenylenediamine, has been used to illustrate the applicability of the technique to determine
for analyzing ascorbic acid in foodstuffs (93) and pharmaceu- ascorbic acid in ascorbic acid tablets. This procedure had a
tical preparations (94). These approaches consisted of retain- detection limit of 0.4 ng/injection. Flow injection analysis has
ing ascorbic acid on an anionic Sephadex column. After inter- also been used to determine the diffusion coefficient of ascor-
fering substances were eluted, ascorbic acid was oxidized to bic acid (104). ....,.-.
dehydroascorbic acid in situ with p-benzoquinone. Dehy- In constant-current coulometric methods used for deter-
droascorbic acid was eluted from the column and reacted mination of this vitamin, a reagent is generated in situ at an
with 4-nitro-o-phenylenediamine. The absorbance was mea- electrode, which then reacts with ascorbic acid. The amount
sured at 375 nm after removal of excess reagent. The method of current per unit time used to obtain the end point is directly
has a detection limit of 15 u,g ascorbic acid/g of sample, and proportional to the original ascorbic acid concentration. Marsh
dehydroascorbic acid can be quantitated after reduction to and coworkers used electrogenerated Br2 to determine the
ascorbic acid with dimercaptoethanol. vitamin in ascorbic acid tablets (105). The relative standard

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

deviation of this method was 2%.
Electrochemical Methods Moros and coworkers (106) described an automated, iodine-
The electro-oxidation of ascorbic acid at an electrode fol- base coulometric method and found it to be equivalent to the
lows an irreversible, EC type of electrode mechanism (95), USP method. This method can analyze 25 samples/4 h and
involving the loss of 2 e - and 2 H + to form dehydroascorbic only requires 1.5 h of constant operator time. The accuracy
acid. This product reacts rapidly via an irreversible hydration and precision of the method were ± 0.3%. Another iodo-
to form diketogluconic acid. Ascorbic acid has a rapid het- metric, constant-current coulometric method has been pub-
erogenous rate constant with many types of electrodes; there- lished, with accuracies within 0.9% and precision of 0.15%
fore, it is not surprising to find many electroanalytical meth- (107). Other coulometric methods involve constant potential
ods. These procedures are based on dynamic electroanalyt- coulometric analysis of ascorbic acid with I2 (108) and cou-
ical techniques (i.e., measurements of i, /, and/or q) and have lometric determination of the vitamin with electrogenerated
been available for many years (96). octacyanomolybdate (109).
Lindquist (97) and Soderhjelm and Lindquist (98) applied
linear sweep voltammetry at the carbon paste electrode to Enzymatic Methods
determine ascorbic acid in fresh fruits, vegetables, multivi- Spectroscopic.—Earlier enzymatic methods using ascor-
tamins, and beverages (97, 98). Voltammograms were run in bic acid oxidase (EC. 1.10.3.3) relied on isolation of the enzyme
metaphosphoric acid solutions at scan rates of 20 mV/s. The from plant material. Few practical methods for determination
linearity range extended between 10~6 and 10_3M. To deter- of ascorbate have been reported, because the enzyme was
mine ascorbic acid in the presence of iron(II), it was neces- not commercially available. Lee and Dawson (110) presented
sary to add 5-nitro-l,10-phenanthroline. This reagent com- an in-depth report on the isolation and characterization of
plexed the iron(II), and the complex had an increased oxi- ascorbate oxidase. An enzymatic procedure for quantitation
dation potential, thus allowing ascorbic acid to be selectively of ascorbic acid in fruit juices has been reported (111) and is
measured. The method was compared with a 2,6-dichloroin- based on the difference in absorbance before and after incu-
dophenol titrimetric procedure and yielded lower results due bation. The method response was linear between 1 and 10
to better selectivity. fxg/mL, with an average recovery of 98 ± 0.8%. The optimal
Differential pulse voltammetry at the glassy carbon elec- experimental parameters for application of this procedure to
trode has been applied to the determination of ascorbic acid plant and animal foods are pH 5.5 for sensitivity and enzyme
in multivitamins containing iron(II) (99). This method did not stability and 30 min incubation at 30°C (112). This procedure
require a complexation agent. The multivitamin preparations gave results comparable to those using 2,6-dichloroindo-
were dissolved in aqueous solution and filtered. This solution phenol or 2,4-dinitrophenylhydrazine.
was then diluted with pH 4 Mcllvaine buffer and differential Recently, Liu and coworkers have used commercially
pulse voltammograms were obtained by scanning at 5 mV/s available ascorbic acid oxidase to develop a suitable method
with a pulse modulation of 50 mV. Measurements of peak for clinical serum or plasma samples (113). In this procedure,
potential for ascorbic acid were obtained at 230 mV versus ascorbic acid is specifically quantitated by pretreating one of
SCE, and the linearity of the method ranged between 10~6 a pair of replicate samples with enzyme. Both samples are
and 10~'M. Differential pulse voltammetry has also been used reacted with Fe(III) and the complexometric reagent 2,4,6-
to determine ascorbic acid in vivo from the striatum of rats tris(2 pyridyl)-5-triazine. The difference in the absorbance
and guinea pigs at an electrochemically pretreated electrode (593 nm) of these samples is proportional to the original
(100). A graphite/epoxy electrode was able to differentiate ascorbic acid concentration. The method was linear between
between ascorbic acid and dopamine (101). This procedure 0.01 and 0.10 mg/mL and had a precision of 2.8%. The major
has a linearity range for ascorbic acid of 5 x 10 -4 and 2 x advantages of the method were (1) endogenous compounds
10~3M in the presence of 2 x 10~5M dopamine. present in biological samples do not interfere, (2) deprotein-
Mason and coworkers have used amperometric detection ization of samples is unnecesary, and (3) the procedure is
of ascorbic acid at a tubular graphite electrode in a flow complete within 25 min with routine clinical laboratory equip-
system (102). The limiting current was monitored at 0.75 V ment.
versus SCE and was linear over the concentration range of The enzyme horseradish peroxidase has also been used for
10"6 to 10~3M. The method is suitable for determining ascor- the determination of ascorbic acid. Roe and Bruemmor have
bic acid in multivitamin preparations and is more specific described a kinetic procedure based on the lag time produced
thaniodometric titration. Although the method involves man- between ascorbic acid,/?-phenylenediamine, and horseradish
ual injection of the samples into a flowing stream (25-30 peroxidase (114). In this procedure, the enzyme converts p-
8 . PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)

phenylenediamine to its colored oxidized form. Ascorbic acid Mapson and Partridge proposed a method to separate
rapidly reacts with the oxidized form, converting it back to ascorbic and isoascorbic acids, using Whatman No. 1 filter
phenylenediamine. The time required for production of a paper with an acidified KCN (potassium cyanide) developing
stable color (of oxidized p-phenylenediamine) is proportional solvent (122). Phenol-acetic acid has been used in combina-
to the original amount of ascorbic acid. This method has been tion with metaphosphoric acid-impregnated paper (123) or
applied successfully to the quality assurance for the labeling glass fiber filter paper-impregnated with silica gel (124) to
compliance of fruit juices. Another variation substitutes eliminate the hazardous solvent combination. The method
guaiacol or homovanilic acid for /?-phenylenediamine and is reported by Mitchell and Patterson (123) was used to quan-
suitable for trace determination in plant extracts (115). titate human urinary ascorbic acid excretion (125). In this
Electrochemical.—Many workers used electrochemical method, ascorbic and erythorbic acids were measured using
techniques to monitor the depletion of oxygen during the a 2-6-dichloroindophenol method after isolating each com-
folio wing. reaction. ponent from the filter paper. Recoveries of 101% and 108%
were reported for rat urine and cress seedlings (126). This
«,„^ Ascorbate' xx' . TT _ procedure involved a preliminary separation using a butanolic
AA + 1/2 Oj'-r-r; DHAA + H 2 0 solvent followed by a second separation with a phenolic sol-
Oxidase vent. Ascorbic acid was eluted and measured via a DCIP

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

Marchesini and coworkers used a Clark electrode to mon- procedure with a recovery of 85% at the 0.5 mg/mL level. A
itor the depletion of oxygen for this reaction (116). Their 2,4-diphenylhydrazine method has also been reported using
procedure was applied to ascorbic acid analysis in fresh and separation by paper chromatography (127).
canned spinach. Dehydroascorbic acid was quantitatable only
after reduction to ascorbic acid with homocysteine, and the Thin layer chromatography has been more useful than paper
method gave results comparable to the 2,6-dichloroindo- chromatography for quantitation of ascorbic acid. For exam-
ple, Saari and coworkers studied oxidation of ascorbic acid
phenol and 2-4 dinitrophenylhydrazine procedures. Posadka
using a 14C-labeled compound and densitometric and radioau-
and Macholan coated a Clark oxygen electrode with a thin
tographic detection (128). The separation system resolved
coat of insolubilized ascorbate oxidase (117). This produced
ascorbic and dehydroascorbic acids and an additional com-
linear results between 10 and 100 mg. Of 35 compounds
pound that was presumed to be 2,3-diketogulonic acid. A
tested, only chlorogenic acid interfered. Ascorbate oxidase
densitometric detection method has also been used to mea-
has also been immobilized on collagen and mounted on a sure ascorbic acid in pharmaceutical syrup formulations con-
Clark electrode (118). The linearity of the method ranged taining vitamins A, Bi, C, and D3 (129). Lyle and Tehrani
from 5 x 10~5 and 5 x 10~4M, and the precision was better (130) described another exotic thin layer chromatographic
than 2.3% over 35 successive assays. The ascorbate-selective method that involves the pyrolysis of thin layer zones directly
electrode is stable for 3 weeks and has been applied to food into a flame ionization gas chromatograph. This method was
analysis. linear over the range of 0.2-3.0 u,g ascorbic acid.
An excellent paper has described coupling ascorbate oxi-
dase with chronoamperometry (119) to determine ascorbic Thin layer chromatographic determinations of ascorbic acid
should attract renewed interest in the coming years because
acid in brain homogenate. In this procedure, a chronoampero-
of advances in high performance thin layer chromatography
metric response for total oxidizable compounds in brain
(131). The major advantages of this separative technique include
homogenate was first determined. A known amount of ascor-
increased sensitivity and resolution, ability to run several
bate oxidase was added, and the total of oxidizable com-
chromatograms in parallel, and capability of using densito-
pounds (minus ascorbic acid) was monitored by chronoam-
metric and fluorimetric detection.
perometry. Ascorbic acid content was determined as the
difference between the 2 readings. The differential technique Gas Chromatography.—-The discovery and.application of
gas chromatography to trace organic analysis produced dra-
requires 1 s or less for each measurement, and the entire
matic improvements in sensitivity and selectivity. The ther-
assay is completed in a few minutes. The method's specificity
mal conductivity and flame ionization detectors were the
has been corroborated by liquid chromatography, and the
most universal detectors/Unfortunately, polar compounds
method has a precision of 2.6%.
(e.g., ascorbic acid) were not easily analyzed by this tech-
Chromatographic Methods nique.
Because many of the spectroscopic and electrochemical Sweeley and coworkers (132) reported the first gas chro-
methods that have been discussed cannot distinguish ascorbic matographic analysis of ascorbic acid. Trimethylsilylether
acid, dehydroascorbic (DHAA), d-isoascorbic (I A A) (also derivatives were synthesized to increase volativity; The deri-
known as erythorbic acid), or other oxidizable compounds, vatization reaction was carried out in a mixture of hexame-
it is not unusual that separation methods have been used to thyldisilazane and trimethylchlorosilane (2 + 1) in pyridine.
increase selectivity. The following chromatographic tech- This basic procedure has been modified for the derivatization
niques will be discussed in some detail: paper and thin layer of ascorbic acid (133-136). The rate of derivatization and
chromatography, gas chromatography, and liquid chroma- elimination of Voluminous ammonium chloride precipitation
tography. has been investigated.; N-Trimethylsilylacetamide (137) or
Paper and Thin Layer Chromatography.—Numerous paper 7,0-bis(trimethylsilyl)acetamide (138) were appropriate
chromatographic methods have been reported for the quali- reagents. Because ascorbic acid may be derivatized at any
tative and quantitative determination of ascorbic acid. De one of its 4 hydroxyl groups, Vecchi and Kaiser (137) char-
Ritter has reviewed many of these procedures (120). Initial acterized the silyl derivative of ascorbic acid with gas chro-
paper chromatographic methods were concerned with iden- matography-mass spectrometry and concluded that the tetra-
tifying suitable visualizing reagents or optimal elution sol- trimethylsilylether derivative was formed.
vents. Partridge reported the first qualitative separation of A GC procedure for quantitation of ascorbic acid, pyridox-
ascorbic and dehydroascorbic acids (121). Cold ammoniacal ine, and nicotinamide in multivitamin preparations has been
AgN0 3 was used to visualize the chromatographic zone. reported by Sennello and Argoudelis (138). This method gave
Dehydroascorbic acid was detected after heating. results similar to the USP procedure and has a linearity range
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. .68, NO. 1, 1985) 9

between 50 and 400 jig and a recoveryof 98%. Another GC An ion-pair LCEC method was used to measure ascorbic
procedure has a linearity range of 0.7-7.0 mg ascorbic acid and erythorbic acids in orange juice, rat plasma, urine, liver,
in orange drink (135). Schlack has also published a method and brain tissues (152). Numerous counter ions were inves-
suitable for food products and pharmaceuticals (136); ascor- tigated, and decylamine gave the best separation. The method
bic acid was precipitated as a lead salt, derivatized, and was highly specific, and no interferences were observed.
detected via flame ionization. Linearity was observed between (b) Spectroscopic detection: Ultraviolet detection has been
0.1 and 1:0 mg, and recovery was 100.1%. used to monitor ascorbic acid after separation by ion-exchange
Liquid Chromatography.—Liquid chromatography (LC) has chromatography. Williams and coworkers (153) determined
been used extensively for ascorbic acid analysis. This tech- ascorbic acid at 254 nm in orange juice, and the method was
nique combines high selectively and sensitivity with rapid specific for ascorbic acid. Another method has described the
sample analysis and does not require derivatization. A mini- quantitation of ascorbic acid in small volumes of aqueous
review of LC methods used for ascorbic acid has recently humor (154). Aqueous humor (2 u,L) was diluted 5-fold, directly
appeared (139). injected onto an ion-exchange column, and detected at 254
(a) Electrochemical detection: The early LC methods for nm. This method gave results comparable to the Roe and
ascorbic acid analysis used electrochemical detection (LCEC). Kuether method. Liebes and coworkers (155) reported the
Kissinger et al. (140) used a thin layer amperometric cell to

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

analyses for ascorbic acid in human lymphocyte extracts.
measure ascorbic acid in urine. Diluted urine was directly Their method also agreed with that of Roe and Kuether.
injected onto a strong anion-exchange column, and ascorbic Floridi and coworkers (156) described a method for deter-
acid was detected at a carbon paste electrode set at + 0.8 V mining ascorbic acid in food products. Linearity ranged from
vs Ag/AgCl. Analyses were complete within 4 min, and high 5 to 30 |xg/mL, and analyses were completed in 20 min.
selectivity, reduced sample preparation, and decreased anal- Recently, Ashoor and coworkers (157) desribed a method for
ysis time were reported. Multivitamin preparations, fruit juices, fruit juices and vegetables.
milk products, and urine were analyzed using this method
(141,142). Samples were extracted with 3% metaphosphoric Reverse phase LC was used to investigate the urinary
acid in 8% acetic acid, and the supernate was diluted with excretion profile of ascorbic acid in human subjects (158).
cold perchloric acid (0.05M) before injection. Liquid chro- These results agreed with the USP titrimetric method. Ascor-
matography with amperometric detection has also been used bic acid was determined in citrus juices with a similar method
to measure ascorbic acid in human serum, plasma, and leu- that gave results comparable to the DCIP method (159). Lin-
cocytes (143). Samples were deproteinized with 6% trichlo- earity was observed from 50, to 400 |xg/mL. The analysis of
roacetic acid, and ascorbic acid concentrations were quan- the vitamin'in a variety of fruit juices and vegetables has been
: titatable between 2 and 30 |xg/mL. A similar method has been described by Watada (160). Recoveries of > 96% were reported,
I reported for urine and plasma (144). and the method was linear up to 7 |xg ascorbic acid.
f. Thrivikraman and coworkers (145) reported a method Sood et al. (161) described the first ionrpair LC method for
I applicable to mouse, rat, and guinea pig tissues. Sample prep-' determining ascorbic acid in foods and multivitamin prod-
I aration consisted of homogenizing the tissue, dilution of the ucts. Several pairing agents were evaluated, and tridecylam-
I supernate, and then separation on an ion-exchange column. monium formate (1 mM) provided the best results. Quanti-
1, Recovery was 102% with a linear range of 15-50 ng/injection. tation at 254 nm gave detection limits of 5 jxg/mL, and their
K This method provides increased sensitivity and selectivity method was in excellent agreement with the USP method.
\ when compared with ultraviolet detection. Low nanogram The method has been modified and used to measure ascorbic
i levels of ascorbic acid have also been monitored in various acid levels in potatoes and potato products (162). These authors
vertebrate and invertebrate tissues using LCEC (146). This reported detection limits of 0.02-1.125 |xg. A method suitable
:
method was compared with the DNPH and a,a'-dipyridyl for the routine analysis of bulk pharmaceutical vitamin-min-
methods (147). The LCEC and DNPH methods compared eral formulations has recently appeared (163). Ascorbic acid
favorably, whereas the a,a '-dipyridyl method gave signifi- levels of 500-5000 (Jig/mL extract were measurable. The abso-
' cantly higher values for many sample types. lute purity of ascorbic acid standards has been determined
Ascorbic acid concentrations in human serum, plasma, and by reverse phase ion-pair LC coupled with conventional stoi-
leukocytes have been determined by using reverse phase ion- chiometric analyses (164). Plots of peak height versus the
pair LC with amperometric detection (148). Samples were volume of standardized iodine solution added were linear.
'• deproteinized before analysis, and 97-100% recoveries were The titration end point was determined by extrapolation in
reported over the range 0.1-5 |xg/mL. Many common drugs zero peak height. Sample purity was calculated by using
• and metabolites did not interfere in this assay. conventional titrimetric equations. This method offers
Green and Perlman (149) and Grun and Loewus (150) have enhanced selectivity and sensitivity as well as reduced sample
used cleanup procedures to improve LCEC selectivity. Plasma requirements over conventional stoichiometric methods.
. ultrafiltration was used to remove protein before separation Bui-Nguyen (165) first reported the separation of ascorbic
by reverse phase ion-pair LC (149). Recovery was 100%, and and isoascorbic acids by LC that used a weak anion-exchange
advantages include increased sensitivity (due to reduced sam- column with detection at 265 nm. These acids were detected
ple dilution) and increased column life. Grun and Loewus in various fruit juices.with a similar method (166), and detec-
(150) used a C-18 Sep-Pak® cartridge to selectively retain tion limits of 20 ng for each acid were reported. Arakowa and
interfering species from extracts of algae and growth media. coworkers (167) have also described a procedure for analyz-
Recovery was >99%. Various sugars, sugar acids, and lac- ing the acids. Ascorbic acid levels of 500 ng/mL were meas-
tones did not interfere. ured at 254 nm, with a total analysis time of 15 min. DH AA
Stillman and Ma (151) studied the use of polarography for and DHIAA could also be indirectly measured following
the quantitation of ascorbic acid in multivitamin preparations. reduction with H2S. This method was used to measure ascor-
Vitamins B,, BJ2, C, and K3 were simultaneously measured bic acid and IAA in rat liver, adrenals, spleen, heart, and
at +1.5 V. Analyses were complete in 9 min; however, no kidney tissues (168). Tissues were homogenized and the
detection limits were reported. supernate was diluted and injected onto an LC column. No
10 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)

interferences were observed, and recoveries of 108% and provided detection limits of about 35 ng total ascorbic acid/
107% were reported for the 2 acids, respectively. mL blood.
The simultaneous determination of dehydroascorbic and
ascorbic acids has been unsuccessful, because DHAA is elec- Future Directions
trochemically inactive or does not absorb UV light at the The search for a single ideal method to analyze ascorbic
same wavelengths as ascorbic acid. Nevertheless, several acid in food products and biological and pharmaceutical sam-
approaches for simultaneously determining ascorbic acid and ples may never be successful. Because each sample type has
DHAA have been reported. A combination of UV and refrac- its own special requirements, a multitude of analytical pro-
tive index detectors has been used to measure these acids in cedures will always appear in the literature. Future methods,
orange juice and urine (169). Isoascorbic, dehydroascorbic, however, will always rely on the basic chemical properties
diketogulonic, and diketogluconic acids were also detected of ascorbic acid or on sophistication of instrumentation to
with this method. A simple procedure that uses dual UV develop selective, sensitive methods. In addition to the many
detectors to quantitate ascorbic acid and DHAA in food (170), suggestions offered in the text, the following avenues for
biological and pharmaceutical samples (171), or guinea pig analysis are offered:
tissue has been reported (172). Filtered liquid samples or Many spectroscopic or electrochemical methods may not,
aqueous extracts of solid samples were directly injected and

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

be selective for ascorbic acid, so future advances may involve
ascorbic acid and DHAA were detected at 254 nm and 210 incorporating solid phase extraction columns (e.g., Bond
nm, respectively. This method was modified to determine Elute®, Sep-Pak® cartridges) to isolate ascorbic acid. This
ascorbic acid and dehydroascorbic acids in fruits and vege- type of isolation is amenable to robotic automation. Robotics
tables (173). Detection limits were 50 ng and 100 ng, and may prove useful in totally automating the semiautomated
recoveries of 93% and 92% were reported for the 2 acids, quinoxaline method of Egberg (87). Future liquid chromato-
respectively. Both compounds have also been determined in graphic methods may rely on using < 3 fim particle size
guinea pig, monkey, and human plasma (174) and orange juice stationary phases or microbore columns for the selective,
(175) with dual UV detector methods. In both procedures, sensitive determinations of ascorbic acid. LCEC methods
the selectivity of the quinoxaline derivatization reaction was will rely on chemically modified electrodes and on advances
used to enhance the UV absorptivity of DHAA. Ascorbic in stationary phases to further improve selectivity and sen-
acid was quantitated by electrochemical (174) and ultraviolet
sitivity.
(175) detection in these procedures.
"Fortified with vitamin C
Dennison and coworkers (176) reported a single wave-
it's a sign we often see
length method in which ascorbic acid was determined directly,
and DHAA was estimated as the difference between ascorbic is it there just like they say
acid and total ascorbic acid at 244 nm. Total ascorbic acid or has it oxidized away." (ref. 141)
was determined following quantitative reduction of DHAA A sample may be scrutinized
with D,L-homocysteine. The acids were measured in a variety to see if the vitamin has oxidized
of beverages, and analyses were made over the range 10-100 no matter what the method.
[xg/mL. Luminescence detection has been investigated for The analyst will detect it.
the direct determination of these acids in ascorbic acid tablets However, the method is only terrific
(177, 178). A lucigenin-base-reductant system was the basis if it's specific.
for the postcolumn reaction of the acids, and conditions were
optimized for maximum ascorbic acid emission. Ascorbic Acknowledgments j
acid and DHAA concentrations of 5-800 (xg/mL and 10-100 The authors gratefully appreciate the assistance of B. A. !
|xg/mL were quantifiable, respectively. DeCastro, D. Gibbons, and D. B. Reynolds.
Derivatization procedures have been used to increase the
sensitivity of LC methods. Garcia-Castineiras and coworkers
REFERENCES [
(179) used the 2,4-dinitrophenylhydrazine method and mon-
(1) Kutsky, J. R. (1973) Handbook of Vitamins and Hormones, j
itored the resulting bis-(dinitrophenylhydrazone) derivative Van Norstrand and Reinhold Co., New York, NY, pp. 71-78 >
at 254 nm. Ascorbic acid concentration was taken as the (2) Hashmi, M. H. (1973) Assay of Vitamins in Parmaceutical j
difference between direct sample derivatization (DHAA and Preparations, J. Wiley and Sons, New York, NY, pp. 286-323 •
DKG) and derivatization following sample oxidation (ascor- (3) Gilpin, R. K., Pachla, L. A., & Ranweiller, J. R. (1981) Anal '
bic acid, DHAA, and DKG). This method was applied to the Chem. 53, 142R-159R
(4) Gilpin, R. K., Pachla, L. A., & Ranweiller, J. R. (1983) Anal.
analysis of aqueous humor samples, and ascorbic acid levels Chem. 55, 70R-87R
of 50-1150 pmol (45-1000 fig/mL sample) were measured. (5) Sauberlich, H. E., Green, M. D., & Omaye, S. T. (1980) in
Advantages of this method included increased specificity and Ascorbic Acid: Chemistry, Metabolism and Uses, P. A. Seib
small sample volume (20 |JLL). Van Boekel and Meeuwissen & B. M. Tolbert (Eds), Advan. Chem. Ser. 200, 199-221 .
(6) McCormick, D. B., & Wright, L. D. (Eds) (1979) Methods in
(180) also used DPNH derivatization to determine ascorbic Enzymology, Academic Press, New York, NY, pp. 62D,
acid in beers. Reaction times were reduced to 2 h by heating Sec. 1
at 70°C, and recovery of the vitamin was 100%. The method (7) Cooke, J. R., & Moxon, R. E. D. (1981) in Vitamin C, J. N.
was highly selective, and no interferences were observed Counsell & D. H. Horning (Eds), Applied Science Publishers,
during the analysis of 27 different beers. Selectivity for intact London, UK, pp. 167-198
(8) Tillmans, J. (1927) Z. Untersuch Lebensm. 54, 33-43
ascorbic acid could be introduced by using the dual sample (9) Karayannis, M. I. (1975) Anal. Chim. Acta 76, 121-130
approach of Garcia-Castineiras (179). Recently, a fluorescent (10) Karayannis, M. I. (1976) Talanta 23, 27-30
assay was used to determine total ascorbic acid in whole (11) Tillmans, J., Hirsch, P., & Jackish, J. (1932) Z. Untersuch.
blood (181). The assay consisted of derivatization of ascorbic Lebensm. 63, 241-267
acid to the quinoxaline derivative after oxidation with ascor- (12) Hoffman, M., Bar-Akiva, A., Tenhum, L., & Berkenstadt, Y.
(1970) Anal. Biochem. 38, 35-39
bic acid oxidase. Recovery from whole blood was 97%, and (13) Egberg, D. C , Larson, P. A., & Honold, G. R. (1973) J. Sci.
fluorescence detection (excitation, 255 nm; emission, 475 nm) Food Agric. 24,789-794
 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985) 11

(14) Rawal, V. M., Patel, U. S., & Desai, R. J. (1978) Ind. J. Med. (61) Schmidt, K., & Holfelder, E. (1975) Z. Lebensm. Unters.
Res. 67, 161-164 Forsch. 157, 217-220
*(15) Nesset, B. L., Windsor, B. L., Humphrey, R. R., &Callantine, (62) Navon, A., & Levinson, H. Z. (1974) Z. Natur. Teil. 29, 777-
M. R. (1967) Anal. Biochem. 19, 89-94 780
(16) King, D. R., & Little, J. W. (1970) 7. Oral. Med. 25, 107-111 (63) Toothill, J., Thompson, S. Y., & Hill, W. B. (1970) 7. Dairy
(17) Cameron, D. J. (1975) Proc. Nutr. Soc. 34, 121 Res. 37, 521-528
(18) Garry, P. J., Owen, G. M., Lashley, D. W., & Ford, P. C. (64) Friberg, U.,Lohmander,S.,&Carlsson,G.(1968)ActaChem.
(1974) Clin. Biochem. 7, 131-145 Scand. 22,3037-3039
(19) Bates, J. F., Hughes, R. E., & Hurley, R. J. (1972) Arch. Oral (65) Hegenauer, J., & Saltman, P. (1972) J. Chromatogr. 74, 133-
Biol. 17, 1017-1020 137
(20) Koch, P., Sidloi, M., & Tonks, D. B. (1980) Clin. Biochem. (66) Aeschbacher, H. V., & Brown, R. G. (1972) Clin. Chem. 18,
13,73-77 965-967
(21) Sauberlich, H., Goad, W. C., Skala, J. H., & Waring, P. P. (67) Attwood, E. C., Robey, E. D., Ross, J., Bradley, F., & Kra-
(1976) Clin. Chem. 22, 105-110 mer, J. J. (1974) Clin. Chim. Acta 54, 95-105
(22) Muller-Molot, W., de Vries, U., Bourgeois, C. F., & Mainguy, (68) Pelletier, O. (1968) J. Lab. Clin. Med. 72, 674-679
P. R. (1975) Z. Lebensm. Unters. Forsch. 159, 47-51 (69) Pelletier, O., & Brassard, R. (1973) Advan. Autom. Anal. 7 3 -
(23) Gorbach, G., Henke, J., & Heical, I. (1968) Mikrochim. Acta 80
886-893 (70) Pelletier, O., & Brassard, R. (1975)7. Assoc. Off. Anal. Chem.
(24) Hiromi, K., Fujimori, H., Yamaguchi-Ito, J., Nakatani, H., 58, 104-109

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

Ohnishi, M., & Tonomura, B. (1977) Chem. Lett. 1333-1336 (71) Pelletier, O., & Brassard, R. (1977)7. Food Sci. 42, 1471-1477
(25) Obata, H., Tokuyama, T., Nitta, Y., Takagi, M., & Hiromi, (72) Behrens, W. A., & Madere, R. (1979) Anal. Biochem. 92, 510-
K. (1979) Agric. Biol. Chem. 43, 2191-2192 516
(26) Hiromi, K., Kuwamoto, C , & Ohnishi, M. (1980) Anal. Biochem. (73) Wahba, N., Yassa, D. A. & Labib, R. D. (1974) Analyst 99,
100, 421-426 397
(27) Koupparis, M. A., Walczak, K. M. & Malmstadt, H. V. (1980) (74) Schmall, M., Pifer, C. W., & Wollish, E. G. (1953) Anal. Chem.
J. Auto. Chem. 2, 66-75 25, 1486-1490
(28) Burger, N., & Kara-Gasparec, V. (1973) Talanta 20, 782-785 (75) Schmall, M., Pifer, C. W., Wolish, E. G., Buschinsky, R., &
(29) Aly, M. M. (1979) Anal. Chim. Acta 106, 379-383 Gainer, H. (1954) Anal. Chem. 26, 1520-1521
(30) Pelizzeti, E., Mentasti, E., & Pramauro, E. (1976) Inorg. Chem. (76) Wilson, S., & Guillan, R. A. (1969) Clin. Chem. 15, 282-291
15, 2898-2900 (77) Smith, J. A. (1972) 7. Endocrinol. 55, 461-462
(31) Pelizzeti, E., Mentasti, E., & Pramauro, E. (1978) Inorg. Chem. (78) Geller, M., Weber, O. W. A., & Senkowski, B. Z. (1969) 7.
17,1181-1186 Pharm. Sci. 58, 477-480
(32) Zannoni, V., Lynch, M., Goldstein, S., & Sato, P. (1974) (79) Crossland, I. (1960) Acta Chem. Scand. 14, 805-813
Biochem. Med. 11, 41-48 (80) Weeks, C. E., & Deutsch, M. (1965) 7. Assoc. Off. Anal.
(33) Okamura, M. (1980) Clin. Chim. Acta 103, 259-268 Chem. 48, 1245-1248
(34) Okamura, M. (1981) Clin. Chim. Acta 115 393-403 . (81) Hashmi, M. H., Shahid, M. A., Akhtar, M. A., & Chughtai,
(35) Lloyd, J. V., Davis, P. S., & Lander, H. (1969) 7. Clin. Path. N. A. (1973) Microchim. Acta 901-906
22, 453-457 (82) Deutsch, M. J., & Weeks, C. E. (1965) 7. Assoc. Off. Agric.
(36) Day, B. R., Williams, D. R., & Marsh, C. A. (1979) Clin. Chem. 48, 1248-1256
Biochem. 12, 22-26 (83) Deutsch, M. J. (1967) 7. Assoc. Off. Anal. Chem. 50, 798-805
(37) Thompsen J. C., & Mottola, H. A. (1984) Anal. Chem. 56, (84) Kirk, J. R., & Ting, N. (1975) 7. Food. Sci. 40, 463-466
755-757 (85) Roy, R. A., Conneta, A., & Saltpeter, J. (1976) 7. Assoc. Off.
(38) Jaselskis, B., & Nelapaty, J. (1972) Anal. Chem. 44, 379-381 Anal. Chem. 59, 1244-1250
(39) Butts, W. C., & Mulvihill, H. J. (1975) Clin. Chem. 21, 1493- (86) Kirk, J. R. (1977) 7. Assoc. Off. Anal. Chem. 60, 1234-1237
1497 (87) Egberg, D. C., Potter, R. H., & Heroff, J. C. (1977) 7. Assoc.
(40) McGown, E. L., Rusnak, M. G., Lewis, C. M., & Tillotson, Off. Anal. Chem. 60, 126-131
J. A. (1982) Anal. Biochem. 119, 55-61 (88) Dunmire, D. L., Reese, J. D., Bryan, R., & Seegers, M. (1973)
(41) Arakawa, N., Tsutsumi, K., Sanceda, N., Kurata, T., & Ina- 7. Assoc. Off. Anal. Chem. 62, 648-652
gaki, C. (1981) Agric. Biol. Chem. 45, 1289-1290 (89) Official Methods of Analysis (1975) 12th Ed., AOAC, Arling-
(42) Shieh, H., & Sweet, T. R. (1979) Anal. Biochem. 96, 1-5 ton, VA, sees 43.051-43.055; 43.056-43.062
(43) Eldaway, M. A., Tawfik, A. S., & Elshabouri, S. R. (1975) (90) DeVries, J. W. (1983) 7. Assoc. Off Anal. Chem. 66, 1371-
Anal. Chem. 47, 461-465 1376
(44) Kamangar, T., Fawzi, A. B., & Maghssoudi, R. H. (1977) 7. (91) Deutsch, M. J. (1984) 7. Assoc. Offic. Anal. Chem. 67, 380
Assoc. Off. Anal. Chem. 60, 528-530 (92) Szepesi, G. (1973) Z. Anal. Chem. 265, 334-336
(45) Hammam, A. S. (1976) 7. Appl. Chem. Biotechnol. 26, 611- (93) Bourgeois, C. F., & Mainguy, P. R. (1974) Int. J. Vit. Nutr.
617 Res. 44, 70-84
(46) Burgess, A. E., & Ottaway, J. M. (1972) Analyst 97, 357-362 (94) Bourgeois, C. F., Czornomaz, A. M., George, P., Belliot, J.
(47) Abdel-Salam, M. A., Abdel-Hamid, M. E., & Abdel-Hady P., Mainguy,. P. R., & Watier, B. (1975) Analusis 3, 540-548
Elsayed, M. (1982) Sci. Pharm. 50, 11-16 (95) Perone, S. P., & Keeflow, W. J. (1966) Anal. Chem. 38, 1760-
(48) Samios, D. N., Karayannis, M. I., & Hadjiioannou, T. P. 1763
(1977) Z. Physik. Chem. Neue. Folge. 104, 189-198 (96) Heyrovsky, J., & Zuman, P. (1968) Practical Polarography,
(49) White, V. R., & Fitzgerald, J. M. (1972) Anal. Chem. 44,1267- Academic Press, New York, NY, p. 102
1269 (97) Lindquist, J. (1975) Analyst 100, 339-348
(50) Hassan, S. S., Abd El Fattah, M. M., & Zaki, M. T. M. (1975) (98) Soderhjelm, P., & Lindquist, J. (1976) Acta Pharm. Suec. 13,
Z. Anal. Chem. 211, 369-371 201-212
(51) Berek, J., & Berka, A. (1975) Anal. Lett. 8, 57-61 (99) Ballantine, J., & Woolfson, A. D. (1980)7. Pharm. Pharmacol.
(52) Kyaw, A. (1978) Clin. Chim. Acta 86, 153-157 32, 353-356
(53) Elnenaey, E. S., & Soliman, R. (1979) Talanta 26, 1164-1166 (100) Buda, M., Gonon, F., Cespuglio, R., Jouvet, M., and Pujol,
(54) Bajaj, K. L., & Kaur, G. (1981) Analyst 106, 117-120 J. F., (1980) C. R. Acad. Sc. Paris 290, 431-434
(55) Jagota, S. K., & Dani, H. M. (1982) Anal. Biochem. 127, 178- (101) Falat, L., & Cheng, H. Y. (1982) Anal. Chem. 54, 2108-2111
182 (102) Mason, W. D., Gardner, T. D., & Stewart, J. J. (1972) 7.
(56) Roe, J. H., & Kuether, C. A. (1943) J. Biol. Chem. 147, 399- Pharm. Sci. 61, 1301-1303
407 (103) Strohl, A. N., & Curran, D. J. (1979) Anal. Chem. 51, 1045-
(57) Zloch, A., Cerven, J., & Ginter, E. (1971) Anal. Biochem. 43, 1049
99-106 (104) Gerhardt, G., & Adams, R. N. (1982) Anal. Chem. 54, 2618-
(58) Zloch, Z. (1975) Mikrochim. Acta I, 213-217 2620
(59) Terada, M., Watanabe, Y., Kunitomo, M., & Hayashi, E. (105) Marsh, D. G., Jacobs, D. L., & Deening, H. (1973)7. Chem.
(1978) An«/. Biochem. 84, 604-608 Ed. 50, 626-628
(60) Beljaars, P. R., Horrocks, W. V. S., & Rondags, T. M. M. (106) Moros, S. A., Hamilton, C. M., Heveran, J. E., Donahue, J.
(1974)7. Assoc. Off. Anal. Chem. 57, 65-67 J., & Oiveri-Vigh, S. (1975) 7. Pharm. Sci. 64, 1229-1231
12 PACHLA ET AL.: J. ASSOC. OFF. ANAL. CHEM. (VOL. 68, NO. 1, 1985)

Edholm. L. E. (1976) Talanta 23, 709-713 146) Carr, R. S., & Neff, J. M. (1980) Anal. Chem. 52, 2428-2430
Karlsson, R. (1975) Talanta 22, 989-993 147) Carr, R. S., Bally, M. B., Thomas, P., & Neff, J. M. (1983)
Cardova-Drellana, R., & Lucena-Conde, X. (1977) Talanta 24, Anal. Chem. 52, 2428-2430
124-125 148) Lee, W., Hamernyik, P., Hutchinson, M., Raisys, V. A., &
Lee, M. H., & Dawson, C. R. (1979) Methods in Enzymologv Labbe, R. F. (1982) Clin. Chem. 28, 2165-2169
62, 30-39 149) Green, D. J., & Perlman, R. L. (1980) Clin. Chem. 26, 796-
Tono, T., & Fujita, S. (1981) Agric. Biol. Chem. 45, 2947-2949 797
Tono, T., & Fujita, S. (1982) Agric. Biol. Chem. 46, 2953-2959 150) Grun, M., & Loewus, F. A. (1983) Anal. Biochem. 130, 191-
Liu, T. Z., Chin, N., Riser, M. D., & Bigler, W. N. (1982) 198
Clin. Chem. 28, 2225-2228 151) Stillman, R., & Ma, T. S. (1974) Mikrochim. Acta 1974, 641-
Roe, B., & Bruemmor, J. H. (1975) Proc. Florida St. Hort 648
Soc. 87,210-213 152) Tsao, C. S., & Salimi, S. L. (1982) 7. Chromatogr. 245, 355-
Celardin, F., Custillo, F. J., &Greppin, H. (1982) J. Biochem. 358
Biophys. Meth. 6, 89-93 153) Williams, R. C , Baker, D. R., & Schmit, J. A. (1973) 7.
Marchesini, A., Montuori, F., Muffato, D., & Maestri, D. Chromatogr. Sci. 11, 618-624
(1974) 7. Food Sci. 39, 568-571 154) Lam, K. W., & Lee, P. F. (1975) Investig. Ophthal. 14, 947-
Posadka, P., & Macholan, L. (1979) Collection Czechoslov 950
Chem. Commun. 44, 3395-3404 155) Liebes, L . . F . , Kuo, S., Krigel, R., Pelle, E., & Silber, R.
Matsumoto, K., Yamada, K., & Osajima, Y. (1981) Anal. (1981) Anal. Biochem. 118, 53-57

Downloaded from https://academic.oup.com/jaoac/article/68/1/1/5704468 by guest on 04 November 2020

Chem. 53, 1974-1979 156) Floridi, A., Coli, R., Fidanza, A. A., Bourgeois, C. F., &
Schenk, J. O., Miller, E., & Adams, R. N. (1982) Anal. Chem. Wiggins, R. A. (1982) Internal. J. Vit. Res. 52, 193-196
54, 1452-1454 157) Ashoor, S. H., Monte, W. C , & Weltey, J. (1984) 7. Assoc.
De Ritter, E. (1965) J. Assoc. Off. Agric. Chem. 48, 985-991 Off. Anal. Chem. 67, 78-80
Partridge, S. M. (1948) Biochem. J. 42, 238-248 158) Wagner, E. S., Lindley, B., & Coffin, R. D. (1979) 7. Chro-
Mapson, L. W., & Partridge, S. M. (1947) Nature (London) matogr. 163, 225-229
164, 479-480 159) Carnevale, J. (1980) Food Chem. Austr. 32, 302-305
Mitchell, L. C , & Patterson, W. I. (1953)7. Assoc. Off. Agric. 160) Watada, A. E. (1982) HortScience 17, 334-335
Chem. 36, 1127-1130 161) Sood, S. P., Sartori, L. E., Wittmer, D. P., & Haney, W. G.
Hornig, D. (1972)7. Chromatogr. 71, 169-170 (1976) Anal. Chem. 48, 796-798
Kadin, H., & Oscada, M. (1959)7. Agric. Food Chem. 7, 358- 162) Augustin, J., Beck, C , & Marousek, G. I. (1981)7. Food Sci.
360 46,312-316
Chen, Y. T., Isherwood, F. A., & Mapson, L. W. (1953) 163) Jenkins, C. (1982) Pharm. Technol. March, 53-65
Biochem. J. 55, 821-823 164) Tsau, J., & Poole, J. W. (1982) Int. J. Pharm. 12, 185-197
Miki, T., Kikuchi, N., & Sahashi, Y. (1962) 7. Vitaminol. 8, 165) Bui-Nguyen, M. H. (1980)7. Chromatogr. 196, 163-165
279-285 166) Geigert, J., Hirano, D. S., & Neidleman, S. L. (1981)7. Chro-
Saari, J. C , Baker, E. M., & Sauberlich, H. E. (1967) Anal. matogr. 206, 396-399
Biochem. 18, 173-177 167) Arakowa, N., Otsuka, M., Kurata, T., & Inagaki, C. (1981)7.
Kouimtzis, Th. A., & Papadazannis, 1. N. (1979) Mikrochim. Nutr. Sci. Vitaminol. 27, 1-7
Acta 1, 145-149 168) Otsuka, M., Kurata, T., Suzuki, E., Arakowa, N., & Inagaki,
Lyle, S. J., & Tehrani, M. S. (1982) Res. Dev. Topics 311 -313 C. (1981) 7. Nutr. Sci. Vitaminol. 27, 9-15
Fenimore, D. C , & Davis, C. M. (1981) Anal Chem. 53, 252A- 169) Doner, L. W., & Hicks, K. B. (1981) Anal. Biochem. 115, 225-
266A 230
Sweeley, C. C , Bentley, R., Makita, M., & Wells, W. W. 170) Finley, J. W., & Duang, E. (1981) 7. Chromatogr. 207, 449-
(1963) 7. Am. Chem. Soc. 85, 2497-2507 453
Dalgliesh, C. E., Horning, E. C , Horning, M. G., Knox, K. 171) Rose, R. C , & Nahrwold, D. L. (1981) Anal. Biochem. 114,
L., & Yarger, K. (1966) Biochem. J. 101, 792-810 140-145
Allison, J. H., & Stewart, M. A. (1971) Anal. Biochem. 43, 172) Rose, R. C , & Nahrwold, D. L. (1982) Anal. Biochem. 123,
401-409 389-393
Koeppen, B. H. (1973) S. Afr. J. Sci. 69, 376-377 173) Wimalasiri, P., & Wills, R. B. H. (1983) 7. Chromatogr. 256,
Schlack, J. E. (1974) 7. Assoc. Off. Agric. Chem. 57, 1346- 368-371
1348 174) Baker, J. K., Kapeghian, J., Verlangieri, A. (1983) 7. Liq.
Vecchi, M., & Kaiser, K. (1967) 7. Chromatogr. 26, 22-29 Chromatogr. 6, 1319-1332
Sennello, L. T., & Argoudelis, C. J. (1969) Anal. Chem, 41, 175) Keating, R. W., & Haddad, P. R. (1982) 7. Chromatogr. 245,
171-173 249-255
Thompson, J. N. (1982) Trace Anal. 2, 1-67 176) Dennison, D. B., Brawley, T. G., & Hunter, G. L. K. (1981)
Kissinger, P. T., Felice, L. J., Riggin, R. M., Pachla, L. A., 7. Agric. Food Chem. 29, 927-929
& Wenke, D. C. (1974) Clin. Chem. 20, 992-997 177) Farber, C. M., Kanengiser, S., Stahl, R., Liebes, L., & Silber,
Pachla, L. A., & Kissinger, P. T. (1976) Anal. Chem. 48, 364- R. (1983) Anal. Biochem. 134, 355-360
367 178) Veasey, R. L., & Nieman, T. A. (1980) 7. Chromatogr. 200,
Pachla, L. A., & Kissinger, P. T. (1979) Methods in Enzymol- 153-162
ogy 62, 15-24 179) Garcia-Castineiras, S., Bonnet, V. D., Figueroa, R., & Miranda,
Tsao, C. S., & Salimi, S. L. (1981) 7. Chromatogr. 224, 477- M. (1981)7. Liq. Chromatogr. 4, 1619-1640
480 180) Van Boekel, M. A. J. S., & Meeuwissen, C. A. J. M. (1983)7.
Mason, W. D., Amick, E. N., & Heft, W. (1980) Anal. Lett. Chromatogr. 261, 176-180
13(B10), 817-824 181) Speek, A. J., Schrijver, J., & Schreurs, W. H. P. (1984) 7.
Thrivikraman, K. V., Refshauge, C , & Adams, R. N. (1974) Chromatogr. 305, 53-60
Life Sci. 15, 1335-1342