Journal of Chromatography A, 1161 (2007) 207–213

Determination of urea using high-performance liquid chromatography with

fluorescence detection after automated derivatisation with xanthydrol
Shona Clark, Paul S. Francis ∗ , Xavier A. Conlan, Neil W. Barnett
School of Life and Environmental Sciences, Deakin University, Geelong, Victoria 3217, Australia
Received 30 April 2007; received in revised form 28 May 2007; accepted 29 May 2007
Available online 2 June 2007

Abstract
A high-performance liquid chromatography (HPLC) method for the determination of urea that incorporates automated derivatisation with xan-
thydrol (9H-xanthen-9-ol) is described. Unlike the classic xanthydrol approach for the determination of urea, which involves the precipitation
of dixanthylurea (N,N -di-9H-xanthen-9-ylurea), the derivatisation procedure employed in this method produces N-9H-xanthen-9-ylurea, which
remains in solution and can be quantified using fluorescence detection (λex = 213 nm; λem = 308 nm) after chromatographic separation from inter-
ferences. The limit of detection for urea was 5 × 10−8 M (0.003 mg L−1 ). This method was applied to the determination of urea in human and
animal urine and in wine.
© 2007 Elsevier B.V. All rights reserved.

Keywords: HPLC; Urea; Xanthydrol; Fluorescence derivatisation; Mass spectrometry

1. Introduction performance liquid chromatography (HPLC) procedures for the

determination of urea [8–12] have not been widely adopted due
Although the published methodology for the determination to their complexity or inadequate sensitivity.
of urea has mainly focussed on clinical applications [1,2], there The determination of urea with xanthydrol (9H-xanthen-9-ol;
is growing demand for sensitive and reliable procedures for Fig. 1a), based on the precipitation of dixanthylurea (N,N -di-
the determination of urea in other matrices, such as milk [3], 9H-xanthen-9-ylurea; Fig. 1b) in concentrated acetic acid, was
soil extracts [4], seawater [5,6] and wine [7]. The most com- first described in the early 20th century [13,14]. The origi-
mon approaches for both research and routine measurement nal procedure involved weighing the dixanthylurea, but other
of urea involve (i) the enzyme (urease) catalysed hydrolysis researchers measured the absorbance of the yellow solution pro-
of urea and subsequent detection of ammonia by colour form- duced when the precipitate was dissolved in 50% sulfuric acid.
ing reactions or electrochemistry or (ii) the direct reaction of Unfortunately, a similar colour was obtained from the unreacted
urea with butane-2,3-dione monoxime (or related reagents) to xanthydrol, which therefore had to be washed from the dixan-
form coloured products. Neither approach is ideal for all appli- thylurea precipitate prior to the addition of the acid [15]. This
cations. The non-enzymatic techniques require noxious reagents approach suffered from poor reproducibility and was unsuit-
and produce an unpleasant odour, which make them unattrac- able for automation. Consequently, it was largely superseded by
tive for clinical and oenological settings [1,7]. However, these procedures based on other colour-forming reactions or urease-
techniques are often favoured for environmental applications catalysed hydrolysis of urea [1,2,16]. Nevertheless, the xanthy-
due to the presence of urease inhibitors and conditions that drol test for urea is still included as part of the AOAC methods
constrain the use of enzymes [4–6]. For many sample types, to confirm the presence of urine on foods and containers [17].
both approaches involve time-consuming pre-treatment steps Xanthydrol also reacts with various primary amides [18–20],
to remove interfering species [4,7]. Previously reported high- indoles [21–23], phenols [19,24] and substituted ureas [19,25],
which has been exploited for quantitative spectrophotometric
analyses [20,23–25]. HPLC procedures with UV absorbance
∗ Corresponding author. Tel.: +61 3 52271294. detection have been developed for the determination of allantoin
E-mail address: psf@deakin.edu.au (P.S. Francis). in plasma extracts [25] and N-acetyl-l-glutamine in rat urine

0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2007.05.085
208 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213

urea and xanthydrol, to improve the analytical application of this

derivatisation reagent.

2. Experimental

2.1. Instrumentation and procedures

2.1.1. High-performance liquid chromatography

Urea was determined using a 1200 series HPLC system with
solvent degasser system, quaternary pump, autosampler and
fluorescence detector (Agilent Technologies, Blackburn, Aus-
tralia). An Eclipse XBD RP-18 column (150 mm × 4.6 mm I.D.,
5 ␮m) from Agilent Technologies was used for all separations
at a constant temperature of 35 ◦ C and the gradients shown in
Table 1. The derivatised analyte was detected with excitation
and emission wavelengths of 213 nm and 308 nm, respectively,
using slit widths of 20 nm.

2.1.2. Automated sample preparation

The HPLC auto-sampler was programmed to perform the
xanthydrol derivatisation (Table 2). The total injection volume
was 40.5 ␮L.

2.1.3. Fluorescence spectra

Fluorescence excitation and emission spectra were collected
with a Cary Eclipse fluorescence spectrophotometer from (Var-
Fig. 1. (a) xanthydrol (9H-xanthen-9-ol); (b) dixanthylurea (N,N -di-9H- ian, Mulgrave, Australia). The selected post-column fraction
xanthen-9-ylurea); (c) N-9H-xanthen-9-ylurea; (d) N2 -acetyl-N-9H-xanthen- from the HPLC procedure was diluted 10-fold with acetonitrile.
9-yl-l-glutamine; (e) xanthylallantoin (N-(2,5-dioxo-4-imidazolidinyl)-N -9H-
xanthen-9-ylurea); (f) xanthylurethane (N-9H-xanthen-9-ylurethane).
Slit widths of 5 nm were used for all experiments. Emission
spectra were corrected for the wavelength dependence of the
detector response and monochromator transmission.
[20]. Both involved manual derivatisation with xanthydrol and
time-consuming steps to remove interfering species prior to
injection [20,25]. For example, as xanthydrol precipitates with Table 1
Mobile phase gradients [solvent A: 20 mM sodium acetate (pH 7.2); solvent B:
urea much faster than N-acetyl-l-glutamine, the dixanthylurea acetonitrile]
was filtered from the reaction mixture approximately 1 h after the
addition of the reagent, and the derivatised analyte was collected Gradient Time (min) Solvent (%B) Flow rate (mL min−1 )
after standing overnight [20]. I 0 20 0.45
Herbert and co-workers reported HPLC procedures for the 0.06 20 0.45
determination of ethyl carbamate (urethane) [26] and urea [27] 12.6 50 0.45
13.6 100 0.45
in alcoholic beverages, with faster manual methods to derivatise 20.6 100 0.8
the analytes. Samples were diluted with ethanol and mixed with 22.6 20 0.45
xanthydrol dissolved in 1-propanol. A small amount of 1.5 M 23.6 20 0.45
hydrochloric acid was then added. Under these conditions, the II 0 20 1.0
analyte was derivatised within 5 min (without the formation of 0.06 20 1.0
precipitates) and could be injected directly onto the column. The 12.6 50 1.0
derivatised analyte was detected by fluorescence (excitation at 13.6 100 1.0
20.6 100 1.0
233 nm and emission at 600 nm) [26,27]. Urea was determined in
22.6 20 1.0
12 port wine samples and although the results appeared reason- 23.6 20 1.0
able, they were not validated against an established procedure
III 0 50 1.0
[27]. 3.3 50 1.0
Based on these findings, we have developed a HPLC pro- 4.0 100 1.0
cedure for the determination of urea incorporating automated 7.0 100 1.0
derivatisation of the analyte with xanthydrol and compared the 7.5 20 1.0
results of this procedure with a conventional approach based on 10.0 20 1.0
10.5 50 1.0
the enzymatic breakdown of urea. Furthermore, we have exam- 12.0 50 1.0
ined the nature of the fluorophore produced in the reaction of
 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213 209

Table 2 ethanol were obtained from Merck (Kilsyth, Australia). Animal

HPLC autosampler program for analyte derivatisation urine samples were supplied by Ray C. Bartolo (Deakin Uni-
Step Action Vial Volume (␮L) versity; Geelong, Australia) and diluted with deionised water
1 Draw 1 14
to within the calibration range (1:5000 dilution for human
2 Draw 2 1.5 urine, 1:2500 for toad, 1:10,000 for rat and 1:100,000 for
3 Mixa 15.5 water deprived mouse. Two Australian wines were purchased
4 Draw 3 0 for analysis: Lindeman Bin 50 Shiraz (2005) and Queen Ade-
5 Draw Sample 25 laide Chardonnay (2005). Prior to derivatisation, all standards
6 Mixa 40.5
7 Draw 3 0
and samples (diluted urine and wine) were mixed with ethanol
8 Wait 5 min (1 mL of standard or sample to 3 mL ethanol) and filtered through
9 Inject 0.45 ␮m HT Tuffryn membranes (Acrodisc PSF syringe filters;
Vial 1: xanthydrol (0.02 M in 1-propanol), vial 2: hydrochloric acid (1.5 M), vial
Pall Australia). Synthetic wine samples consisted of ethanol
3: acetonitrile for needle wash. (12%, v/v), tartaric acid (1200 mg L−1 ), potassium hydro-
a In air; maximum speed; three times. gen tartrate (2110 mg L−1 ) malic acid (5000 mg L−1 ), catechin
(red wine, 1500 mg L−1 ; white wine, 30 mg L−1 ), caffeic acid
(100 mg L−1 ), gallic acid (100 mg L−1 ), quercetin (red wine,
2.1.4. Mass spectra
50 mg L−1 ; white wine, 10 mg L−1 ), proline (200 mg L−1 ),
A 6210 MSDTOF mass spectrometer (Agilent Technologies)
glutamic acid (200 mg L−1 ) and arginine (100 mg L−1 ). The
was used with the following conditions: drying gas, nitrogen
synthetic wine components were obtained from Sigma–Aldrich,
(7 mL min−1 , 350 ◦ C); nebulizer gas, nitrogen (16 psi); capil-
except for ethanol, which was purchased from Merck.
lary voltage, 4.0 kV; vaporizer temperature, 350 ◦ C; and cone
voltage, 60 V.
3. Results and discussion

2.1.5. Urease test kit 3.1. Sample preparation and chromatographic separation
A K-URAMR test kit (Megazyme, Bray, Ireland) was used for
comparison purposes. The test kit method involved urease catal- In the manual pre-column derivatisation method described
ysed hydrolysis of urea to ammonia and the subsequent reaction by Herbert and co-workers, the samples were made up to 70%
of ammonia, 2-oxoglutarate and reduced nicotinamide–adenine ethanol. The diluted samples (500 ␮L) were then combined with
dinucleotide phosphate (NADPH) in the presence of glutamate 100 ␮L of the derivatising reagent (0.02 M xanthydrol in 1-
dehydrogenase to form glutamic acid and NADP+ . The con- propanol) and 50 ␮L of 1.5 M hydrochloric acid [26,27]. After
sumption of NADPH is measured by the decrease in absorbance 5 min, 30 ␮L of the reaction mixture was injected onto a Hypersil
at 340 nm and is proportional to the original amount of urea over ODS column (200 mm × 2.1 mm, 5 ␮m particle size) and sep-
a finite range [28]. Tests were performed as described in the arated using gradient elution (Table 1; Gradient I). In the case
assay procedure. Urine samples were filtered through 0.45 ␮m of urea, the derivatised analyte was eluted within 7 min [27].
HT Tuffryn membranes (Acrodisc PSF syringe filters; Pall Aus- We repeated this procedure using an Eclipse XBD RP-18 col-
tralia, Cheltenham, Australia) and diluted between 2500- and umn (150 mm × 4.6 mm I.D., 5 ␮m particle size) at a constant
100,000-fold. A volume of 1.0 mL of diluted urine was used temperature of 35 ◦ C, and found a urea peak retention time of
for each test. White wine samples were filtered and a volume 14.8 min. By increasing the flow rate and the proportion of ace-
of 0.1 mL was used for each test. Where required, the pH of tonitrile in the mobile phase (Table 1; Gradients II and III), the
the wine sample was adjusted to approximately pH 8.0 using retention time was reduced to 6.0 min (Gradient II) or 2.1 min
2 M NaOH and then made up to double the original sample (Gradient III), without co-elution from the excess xanthydrol.
volume with deionised water. Samples and reagent solutions Overall run times (excluding derivatisation) were 23.6 min using
were dispensed with handheld autopipettes (Gilson Pipetman either Gradient I or II and 12 min using Gradient III.
models P20, P200, P1000; John Morris Scientific, Balwyn, Aus- To simplify the overall procedure, the derivatisation was auto-
tralia) into standard 10 mm disposable cuvettes (Kartell type mated using the HPLC autosampler. The maximum volume of
1939: UV grade, PMMA, 4.5 mL volume; Crown Scientific, solution that can be drawn into the sample loop of the autosam-
Wantirna South, Australia) and mixed by repeated inversion pler is 100 ␮L and therefore the volumes described in the manual
after the cuvettes were sealed with either Para film or cuvette derivatisation method by Herbert and co-workers [26,27] were
lids (Kartell type 1962; Crown Scientific). The absorbance of scaled down by a factor of 20. Under these conditions, the peak
the solutions was measured at 340 nm using a Cary 300 Bio height and area for the derivatised urea using the automated
UV-visible spectrophotometer (Varian). method were slightly lower than when the manual procedure
was used. The ratio of reagent and sample solutions was then
2.2. Reagent and samples optimised to gain the highest response. The final settings for
the auto-sampler program are shown in Table 2. The repro-
Urea and xanthydrol were purchased from Sigma–Aldrich ducibility of the manual and automated methods was compared
(Castle Hill, Australia). Sodium acetate was obtained from using 10 replicate injections of a urea standard (1 × 10−4 M).
BDH (Poole, UK). Hydrochloric acid, propanol, acetonitrile, The automated derivation was more precise (RSD of 1.8%) than
210 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213

Table 3
Analytical figures of merit for the determination of urea using HPLC with automated analyte derivatisation

Gradient Retention time (min) LODa (M) LOQb (M) Equation (log–log)c R2 RSDd (%)

I 15.0 5 × 10−8 2.5 × 10−7 y = 1.04x + 7.14 0.9991 0.75

II 6.0 5 × 10−8 2.5 × 10−7 y = 1.02x + 7.09 0.9995 1.26
III 2.1 5 × 10−8 2.5 × 10−7 y = 1.02x + 7.13 0.9994 1.90
a Limit of detection (blank + 3σ).
b Limit of quantification (blank + 10σ).
c Between 1 × 10−7 and 1 × 10−3 M.
d Relative standard deviation (n = 10) for peak area.

the manual procedure (RSD of 2.6%), although both were at an the xanthydrol derivative of ethyl carbamate, which is likely to
acceptable level. The analytical figures of merit for the HPLC be xanthylurethane (Fig. 1f), has similar fluorescence character
method with automated derivatisation (and the three gradients [26,27] to the product under investigation.
described above) are shown in Table 3. The limit of detec- The formation of N-9H-xanthen-9-ylurea in the derivatisa-
tion for urea of 5 × 10−8 M (0.003 mg L−1 ) was superior to the tion procedure was confirmed by examining the positive ion
concentrations reported in previous HPLC procedures for the mass spectrum of the relevant eluent fraction. The characteristic
determination of urea [4,7]. N-9H-xanthen-9-ylurea [M + H]+ ion at m/z 241 and a fragment
ion at m/z 196 were both observed. The base peak in the mass
3.2. Fluorophore characterisation spectrum was at m/z 181, which corresponds to the resonance sta-
bilised xanthenyl ion. Characteristic ions for dixanthylurea were
To examine the nature of the derivatised analyte, the rele- also present (m/z 423 for [M + H]+ and m/z 239 for the parent
vant fraction of the column eluent (corresponding to a retention compound minus a xanthenyl radical [M − 181]+ [29]), but these
time range of 14.5–15.5 min using Gradient I) was collected and peaks were relatively low in intensity (∼1%) in comparison to
examined using a spectrofluorometer and mass spectrometer. For the ions that were representative of the N-9H-xanthen-9-ylurea.
the fluorescence detection of the derivatised urea, Herbert and
co-workers [27] used excitation and emission wavelengths of 3.3. Preliminary evaluation with real samples
233 nm and 600 nm, respectively. Using a spectrofluorometer,
we observed similar maxima at 211 nm and 240 nm in the exci- The proposed HPLC procedure with automated derivatisa-
tation spectrum and 592 nm in the emission spectrum. However, tion for the determination of urea was applied to urine and wine
a more intense peak at 298 nm was detected in the emission spec- samples. The determination of urea in urine is commonly used
trum. Furthermore, both peaks (298 nm and 592 nm) remained in clinical laboratories to assess the nitrogen balance of hospi-
in the spectrum when an emission filter that only transmitted talised patients on specialised nutritional support regimens [30]
light between 250 nm and 395 nm was introduced and neither and is important for research into the mechanism of urinary con-
peak was detected when an emission filter that transmitted light centrating ability and renal urea handling in mammals [31]. The
between 550 nm and 1100 nm was used. It was therefore con- determination of urea in wine has been used to establish the
cluded that the apparent peak at 592 nm was actually an artefact potential for ethyl carbamate formation. The concentration of
of the instrumentation (i.e. second-order diffraction with respect ethyl carbamate (a group 2A carcinogen; IARC, 2007) in alco-
to the peak at 298 nm, arising from the grating within the holic beverages is regulated or monitored by organisations such
spectrofluorometer). The excitation and emission wavelengths as the Canadian Health Protection Branch and the U.S. Food and
were re-evaluated using the HPLC fluorescence detector and the Drug Administration [7,32].
optimum values were found to be 213 nm and 308 nm, respec- Human and animal urine samples were prepared by dilution
tively. The slight discrepancies between these values and those with deionised water to within the calibration range. For the
obtained with the spectrofluorometer were due (in part) to the HPLC procedure, 1 mL of each diluted urine sample was mixed
different slit widths used in each instrument. with 3 mL of ethanol. Typical chromatograms for human urine
The classic determination of urea by derivatisation with xan- samples using mobile phase Gradients I and III (Table 1) are
thydrol in concentrated acetic acid is based on the formation shown in Fig. 2. In both cases, the peak for the derivatised analyte
of dixanthylurea (Fig. 1b), which precipitates from the reaction was resolved from all interferences. The results from the HPLC
mixture [15]. Coxon et al. [29] treated urea with xanthydrol procedure were compared to those obtained using a commer-
under less acidic conditions (between 0.5 and 1.0 M acetic acid) cially available manual test kit, based on the enzyme-catalysed
to produce N-9H-xanthen-9-ylurea (Fig. 1c), which was more hydrolysis of urea. Each test kit assay was performed using a
soluble in polar solvents than dixanthylurea. The derivatisation diluted urine sample volume of 1.0 mL. As shown in Table 4,
described in this paper was performed in propanol and rela- the results for the two procedures were in good agreement.
tively dilute hydrochloric acid and does not involve precipitation, The concentration of urea in wine is far lower than the levels
which suggests that N-9H-xanthen-9-ylurea is the dominant found in urine and therefore sample preparation for the deter-
reaction product. Other monoxanthyl derivatives (see Fig. 1d mination of urea in wine using the HPLC procedure involved
and e) have been prepared for HPLC procedures [20,25] and only dilution with ethanol (to avoid precipitation) and filtration.
 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213 211

Fig. 2. Determination of urea in human urine using (a) Gradient I and (b)
Gradient III. Fig. 3. Determination of urea in white wine using (a) Gradient I and (b) Gradient
II.

Gradient III could not be used due to co-eluting interferences.

Typical chromatograms for the analysis of white and red wines, when Gradient II was used (Fig. 5), which had a greater effect
using Gradients I and II are shown in Figs. 3 and 4. Samples on the analysis of red wine due to the higher concentration of
were spiked with urea to confirm the retention time in the wine flavonoids such as catechin. Consequently, Gradient I was used
matrix. For white wine samples, the peak for urea appeared to for all subsequent analysis of wines. Wine can contain relatively
be sufficiently resolved using either gradient (Fig. 3). There was high concentrations of ammonia compared to urea. Therefore, a
however a more noticeable rise in the baseline when Gradient 100 ␮L sample volume of undiluted wine was used in the test kit
II was used. For red wine samples, only Gradient I provided procedure to place the total amount of ammonia (both endoge-
sufficient separation of the derivatised urea from interferences nous and the product of urea hydrolysis) within the linear range.
(Fig. 4). Furthermore, an examination of the main components However, this reduction in sample volume raises the theoretical
of wine revealed that catechin co-eluted with the target analyte minimum error margin for the determination of the urea (cor-
responding to the smallest differentiating absorbance of 0.005
Table 4 units) to 0.62 mg L−1 .
Urea in human and animal urine determined using the HPLC procedure with To examine the accuracy of the proposed approach, a syn-
xanthydrol derivatisation and a commercially available test kit thetic wine sample was spiked with six different concentrations
Concentration of urea (M) of urea and analysed with both the HPLC and test kit proce-
HPLCa HPLCb Test kit
dures (Table 5). The percentage recoveries were close to 100%
using the HPLC procedure, but were lower than 100% using
Human 1 0.28 0.29 0.29 the test kit when the concentration of urea was below 6 mg L−1 .
Human 2 0.47 0.48 0.49
Mouse 1 1.49 1.64 1.30
The greatest difference occurred at the lowest urea concentration
Toad 1 0.078 0.081 0.080 (2.0 mg L−1 ), at which the result for the test kit was 1.5 mg L−1 .
Toad 2 0.079 0.082 0.079 The addition of ammonium chloride (40 mg L−1 ) to the syn-
Rat 1 0.29 0.30 0.30 thetic wine matrix accentuated this discrepancy by lowering the
a Gradient I. test kit result for urea to 1.0 mg L−1 . The HPLC procedure was
b Gradient III. unaffected by the presence of ammonium chloride.
212 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213

Fig. 5. Separation of urea and catechin in synthetic wine using (a) Gradient I
and (b) Gradient II.
Fig. 4. Determination of urea in red wine using (a) Gradient I and (b) Gradient
II.
Table 6
Urea in spiked white wine determined using the HPLC procedure with xanthy-
The two procedures were then applied to the determination drol derivatisation and a commercially available test kit
of urea in an Australian white wine and five samples of the wine
spiked with urea (Table 6). For each sample, the concentration Urea (mg L−1 ) Recovery (%)
of urea determined using the HPLC procedure was higher than Spike HPLC Test kit HPLC Test kit
that of the test kit. Overall, the deviation from 100% recovery for 0 2.0 0.9 – –
the test kit was greater than the deviation for the HPLC proce- 1 3.0 2.0 100 110
dure. However, the percentage recovery for the HPLC procedure 2 3.9 2.6 95 85
declined (in general) as the concentration of urea was increased, 4 5.9 4.5 98 90
which is indicative of matrix effects. These effects may arise due 8 9.6 9.1 95 103
10 11.2 11.1 92 102
to the consumption of the reagent by other sample components,
which reduce the rate that the target analyte is derivatised. When
the standard addition method was performed, the concentration
of urea in the white wine was found to be 2.2 mg L−1 , compared
to 2.0 mg L−1 using the external standard approach.
Table 5
Urea in synthetic white wine determined using the HPLC procedure with xan-
thydrol derivatisation and a commercially available test kit 4. Conclusion
Urea (mg L−1 ) Recovery (%)
The HPLC procedure described in this paper is a simple,
Actual HPLC Test kit HPLC Test kit sensitive, low-cost alternative for the determination of urea in
2.0 2.0 1.5 100 75 complex sample matrices. The automation of the xanthydrol
3.0 3.0 2.6 100 87 derivatisation step reduced manual labour, sample preparation
4.0 3.9 3.5 98 88 time and reagent consumption, and improved the reproducibility
6.0 6.3 6.5 105 108 of the analysis. For the determination of urea in human or ani-
10.0 10.1 10.3 101 103
12.0 11.9 12.1 99 101
mal urine, a rapid and robust separation with a retention time for
urea of approximately 2 min was developed. The results were in
 S. Clark et al. / J. Chromatogr. A 1161 (2007) 207–213 213

good agreement with those using a commercially available test [13] R. Fosse, Compt. Rend. 158 (1914) 1432.
kit. The relatively high concentrations of ammonia compared [14] L. Hugounenq, A. Morel, C. R. Seances Soc. Biol. Ses. Fil. 74 (1913)
to urea in wines can lower the accuracy of test kits based on 1055.
[15] R.G. Martinek, J. Am. Med. Technol. 31 (1969) 678.
the enzyme catalysed hydrolysis of urea, but the HPLC proce- [16] A.R. Butler, D. Walsh, Trends Anal. Chem. 1 (1982) 120.
dure was unaffected by the presence of ammonia. Some matrix [17] W. Horwitz (Ed.), Official Methods of Analysis of AOAC International,
effects were observed and therefore calibration using standard 17th ed., Association of Official Analytical Chemists (AOAC) Interna-
additions may be required for the analysis of wine. Further- tional, Gaithersburg, MD, 2000, p. 56, Ch. 16.
more, this approach has great potential for the development of [18] R.F. Phillips, B.M. Pitt, J. Am. Chem. Soc. 65 (1943) 1355.
[19] W. Lang, Arzneim.-Forsch. 1 (1951) 230.
new HPLC methods to determine other compounds that react [20] U. Hellmann, I. Luederwald, M. Neuhaeuser, Fresenius Z. Anal. Chem.
with xanthydrol, such as amides and substituted ureas. 325 (1986) 290.
[21] W.R. Fearon, Analyst 69 (1944) 122.
References [22] S.R. Dickman, W.L. Westcott, J. Biol. Chem. 210 (1954) 481.
[23] S.R. Dickman, A.L. Crockett, J. Biol. Chem. 220 (1956) 957.
[24] S. Mizukami, Y. Kanaya, Yakugaku Zasshi 83 (1963) 1152.
[1] A.J. Taylor, P. Vadgama, Ann. Clin. Biochem. 29 (1992) 245.
[25] K. Hirota, M. Kawase, S. Ohmori, T. Kishie, J. Chromatogr. 277 (1983)
[2] P.S. Francis, S.W. Lewis, K.F. Lim, Trends Anal. Chem. 21 (2002) 389.
165.
[3] D. Lefier, Bull. Int. Dairy Fed. 315 (1996) 35.
[26] P. Herbert, L. Santos, M. Bastos, P. Barros, A. Alves, J. Food Sci. 67 (2002)
[4] D.F. Lambert, J.E. Sherwood, P.S. Francis, Aust. J. Soil Res. 42 (2004)
1616.
709.
[27] M.J. Veiga, P. Herbert, T. Simões, A. Oliveira, A. Alves, Presented at the
[5] N.M. Price, P.J. Harrison, Mar. Biol. 94 (1987) 307.
4th International Conference on Instrumental Methods of Analysis: Modern
[6] M. Revilla, J. Alexander, P.M. Glibert, Limnol. Oceanogr. Methods 3
Trends and Applications, Iraklion, Greece, October 2005, Poster P-III-16.
(2005) 290.
[28] Urea/Ammonia (Rapid) Assay Procedure K-URAMR 11/05, Megazyme
[7] P.S. Francis, Aust. J. Grape Wine Res. 12 (2006) 97.
International, Bray, Ireland, 2004.
[8] R.B.H. Wills, J. Silalahi, M. Wootton, J. Liq. Chromatogr. 10 (1987) 3183.
[29] B. Coxon, A.J. Fatiadi, A. Cohen, H.S. Hertz, R. Schaffer, Org. Magn.
[9] A. Abshahi, S.S. Goyal, D.S. Mikkelsen, Soil Sci. Soc. Am. J. 52 (1988)
Reson. 13 (1980) 187.
969.
[30] R.N. Dickerson, A.C. Tidwell, G. Minard, M.A. Croce, R.O. Brown, Nutri-
[10] T. Matsudo, M. Sasaki, Biosci. Biotechnol. Biochem. 59 (1995) 827.
tion 21 (2005) 332.
[11] S. Kodama, T. Suzuki, J. Food Sci. 60 (1995) 1097.
[31] B. Yang, L. Bankir, Am. J. Physiol. Renal Physiol. 288 (2005) F881.
[12] P. Dallet, L. Labat, E. Kummer, J.P. Dubost, J. Chromatogr. B 742 (2000)
[32] B. Zimmerli, J. Schlatter, Mutat. Res. 259 (1991) 325.
447.