712  Hojjatie & Abrams: Journal of AOAC International Vol. 97, No.

3, 2014

SPECIAL GUEST EDITOR SECTION

Validation for the Determination of Biuret in Water-Soluble,

Urea-Based Commercial Inorganic Fertilizer Materials, Urea
Solutions, and Sulfur-Coated Urea Products by Reversed-Phase
Liquid Chromatography: Single-Laboratory Validation of an
Extension of AOAC Official MethodSM 2003.14
Michael M. Hojjatie and Dean Abrams
Tessenderlo Kerley, Inc., Research and Development Department, 2045 N. Forbes Blvd, #106, Tucson, AZ 85745

A single-laboratory validation (SLV) study for the LC 4.73%. Furthermore, comparative studies for biuret
determination of biuret in dry and liquid urea-based using AOAC Official MethodsSM 960.04 and 976.01
commercial fertilizers was conducted. A total of with the proposed LC method were performed. The
23 samples were used: 11 commercial dry urea three methods produced very close results; however,
products, two urea ammonium nitrate products, the two AOAC methods generate hazardous wastes
eight liquid urea-based commercial fertilizers, and and are more tedious. On the basis of accuracy
four sulfur-coated urea samples from different and precision of the results for this SLV study, it is
sources. In addition, one biuret standard from Aldrich recommended that this method be collaboratively
and one sample from the Magruder check sample studied for the determination of biuret in dry and
program were used as validation samples. The liquid urea-based commercial fertilizer materials.
proposed method is an extension of AOAC Official
MethodSM 2003.14 and is based on dissolution of the

B
test portion in the LC mobile phase and determination iuret is one of several compounds formed when molten
by HPLC. The system is linear over a concentration urea is heated near or above its melting point (132°C
range of 1.00–4.50 mg/L biuret, with a correlation or 270°F) during the manufacturing of urea (1). In
coefficient ≥0.9999. The biuret was well- separated pebbling urea, up to 2% of biuret is formed at 320°F (2). The
from urea in the commercial urea samples, and conversion of urea to biuret is represented by the equation:
from other constituents in the commercial liquid
fertilizer with no observed interferences. Recoveries 2 CO(NH2)2 → NH2CONHCONH2 + NH3
were determined by spiking four of the validation Urea Biuret
materials with a known amount of biuret standard
and measuring the biuret level according to the Urea has become the leading source of nitrogen (N)
method. The averaged recovery was 97%. Method containing fertilizer, and the potential adverse effects of biuret
precision was determined by quadruplicate analyses on plant growth are a concern (3). In high concentrations, biuret
of four of the liquid and six of the commercially applied to soil or plant foliage interferes with N metabolism
available dry urea validation materials using three and hinders protein synthesis in plants. Biuret is mineralized by
and four replicate analyses. For the liquid fertilizer many soil microorganisms, but the process is much slower than
analyses, the RSD ranged from 7.04 to 13.31%. For for urea. Excessively high biuret concentrations can damage
the dry urea analyses, the RSDs ranged from 5.68 to seedlings and, like urea, should not be placed in close proximity
14.34%. Instrument precision was evaluated at the to germinating seeds. Crop tolerance to biuret varies according
test initiation by using seven injections of five biuret to the plant species, soil conditions, fertilizer placement, and
standard solutions. SD varied from 0.27 to 1.02%, method of application. The toxicity of biuret to some plant
with RSDs averaging 1.14%. The LOD was determined species has been used as an argument against the use of urea as a
to be 0.009% biuret in material, while the LOQ was fertilizer, due to the potential for biuret contamination (3). As a
determined to be 0.031% biuret in material. In addition result of crop losses that have been attributed to the application
to the intralaboratory study, interlaboratory studies of urea high in biuret, recommendations against the use of both
were performed by two other outside laboratories solid urea and liquid fertilizer products containing urea have
using this method. Over a concentration range of 0.2 been issued, both for direct application and for ammoniating
to 0.9% biuret, the average SD was 0.11%, the average mixed fertilizers (4).
RSD was 21.16%, and the average HorRat value was A review of foliar literature indicates that crop foliage has a
wide range of tolerance to biuret (5). For example, pineapple can
Guest edited as a special report from the AOAC Agricultural tolerate up to 3% biuret (6), whereas citrus recommendations
Community on “Collaborations in New and Improved Methods of
Analysis for Plant Food Materials” by Nancy Thiex. for urea sprays suggest no more than 0.25% biuret (7). A 3-year
Corresponding author’s e-mail: [email protected] study showed that potato yields were reduced by 5.1, 10.7,
DOI: 10.5740/jaoacint.12-443 13.8, and 18.2% for urea foliar sprays containing 0.5, 1.0, 1.5,
 Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014  713

Table  1.  List of validation materials Table  2.  List of biuret standards used (prepared from
Biuret–1, row 22 in Table 1)
Average biuret, Urea,
No. Type of fertilizer Reported grade % % Stock standard, Standard Volume transferred, Final concn,
a g/100 mL No. mL into 100 mL ppm SDa
1 Liquid–1 28-0-0 (70% SRN) 0.30 11.73
2 Liquid–2 30-0-0 (70% SRN) 0.10 12.33 0.0048 1 2 0.9312 0.0324
2 4 1.8624 0.019
3 Liquid–3 28-0-0 (72% SRN) 0.10 7.30
3 6 2.7936 0.008
4 Liquid–4 30-0-0 (60% SRN) 0.20 11.25
4 8 3.7248 0.027
5 Liquid–5 26-0-0 (33% SRN) 0.23 16.92
5 10 4.6560 0.016
6 Liquid–6 29.5-0-0 (60% SRN) 0.10 8.26
a
  Average SD = 0.078.
7 Liquid–7 30-0-0 (50% SRN) 0.20 13.90

8 Liquid–8 26-0-0-0.5B 0.23 16.11

(33% SRN) These classes of products also contain a substantial component
9 Liquid–9 29.5-0-0 (71% SRN) 0.03 8.65 of urea-formaldehyde reaction products. Due to the presence
10 Liquid–10 28-0-0 (73% SRN) 0.03 6.18 of these co-products, quantitative determination of the claimed
urea content is important. For the reasons discussed above,
11 Urea–1 46-0-0 0.68 46.0
quantitative determination of biuret content of the urea is
12 Urea–2 46-0-0 1.0 46.0 equally important.
13 UAN–1 32-0-0 0.20 16.0 AOAC INTERNATIONAL currently has two methods, Official
SM
14 UAN–2 32-0-0 0.20 16.0 Method   960.04 (spectrophotometric) and AOAC Official
MethodSM 976.01 (atomic absorption spectrophotometric; AAS),
15 Urea–3 46-0-0 0.20 46.0
for determination of biuret in fertilizers (15). AOAC Method 960.04
16 Urea–4 46-0-0 0.70 46.0 is applicable only for urea products and not for mixed fertilizers.
17 Urea–5 46-0-0 0.70 46.0 The method is based upon extraction with water/alcohol,
18 Urea–6 46-0-0 1.0 46.0 complexation of biuret with copper (Cu) solution, and subsequent
determination of the Cu concentration by spectrophotometry.
19 Urea–7 46-0-0 (0.2–08%) 0.20 46.0
Biuret is determined from the Cu concentration. AOAC 976.01
20 Urea–8 46-0-0 (0.2–08%) 0.28 46.0 method is also based on complexation with Cu using CuSO4 and
21 Urea–9 46-0-0 (0.5–1.0%) 0.68 46.0 determination by AAS. Both methods require several reagents,
22 Urea–10 46-0-0 (0.5–1.0%) 0.91 46.0 such as CuSO4, tartrate, alcohol, and indicator solutions. These
methods are complex and laborious, and are therefore more prone
23 Urea–11 46-0-0 (0.5–1.0%) 0.61 46.0
b
to multiple sources of operator error.
24 Biuret–1 NA 97 pure NA Other methods are also mentioned in the literature for
25 Magruder c
NA 0.12 NA determination of biuret. In one literature source (15), a modified
a
SRN = Slow Release N.
 
colorimetric method was used for the determination of urea
b
pyrolyzate products containing biuret over a wide concentration
NA = Not available.
 
range. This method is based on an advantageous equilibrium
c
Magruder check sample 2012-05A.
  existing between copper–tartrate and copper–biuret complexes in
Fehling’s solution.
and 2.0% biuret, respectively. In this study, three 12 lb N/acre Another colorimetric method for the determination of
hydroquinone and a procedure for the analysis of biuret in
applications were made during the growing season using a 2%
hydroquinone-containing urea are described by Xiaoyan
urea spray solution (8).
et al. (16). For hydroquinone determination, NH4Cl (pH 10) was
Biuret, if banded near the seed in excessive amounts, inhibits
used as a color reagent, and the color intensity was measured
seed germination and injures or kills seedlings. When excessive
at 360 nm with a UV-Vis spectrophotometer. The recovery of
biuret was applied to rice, the rice yields were reduced by 21,
hydroquinone by this method was 90–110%.
28, and 34% for urea sprays containing 1, 2, and 3% biuret,
Xiaoyan et al. (16) also reported determination of biuret in
respectively (9). hydroquinone-urea using a copper colorimetric method, but with
The literature on the biuret level for foliar urea applications pre-removal of hydroquinone and its oxides with butyl acetate
indicates that a level of 0.3% or less is required to avoid leaf using a chromatography tube. In a modified version, the tube
injury (2, 3). Of course, the potential for injury is dependent on was filled with aluminum oxide and silica-magnesia absorbent,
such factors as N application rate, N concentration in the spray, respectively. This procedure gave a total of 88.3–93.0% recovery
crop, growth stage, air temperature, and humidity (2). of biuret. Farlin et al. (17) described a method for quantitative
Urea-formaldehyde condensation products, commonly determination of biuret in physiological fluids using ion-exchange
known as ureaforms, have been used for many years as a source chromatography and ninhydrin.
of slow-release N fertilizers for plants (10–14). In commercial Finally, a method has been developed for the determination of
fluid fertilizers derived from reaction of urea and formaldehyde, biuret in urea based on the separation of the Cu-biuret complex
unreacted urea is always a substantial component of such with the aid of an anion exchanger, followed by titration
products and is quantitatively claimed on the registration labels. of the bound Cu with EDTA (18). This method may be used
714  Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014

Table  3.  Instrument precision based on repeated injections of biuret standards

Biuret, g/100 mLa Peak area Mean SD RSD, %

0.0001 30.18 30.50 29.70 30.50 30.25 29.50 32.60 30.46 1.02 3.34
0.0002 60.40 59.90 59.70 60.84 60.73 61.00 61.40 60.57 0.61 1.00

0.0003 89.50 90.10 89.90 89.99 90.20 89.60 90.10 89.91 0.27 0.30

0.0004 118.95 118.64 119.27 117.78 118.70 117.90 116.80 118.29 0.85 0.72

0.0005 147.60 147.60 147.60 148.10 148.40 147.00 147.00 147.61 0.52 0.35
Average 0.65 1.14
a
  Biuret standard is 97% pure. The weight in the standard curve is corrected for 100% pure biuret.

directly for the determination of biuret in solutions and in end- different sources were used. In addition, one biuret standard
products of urea production, even with biuret contents down to from Aldrich Chemical Co. (Milwaukee, WI) and one from
0.001% (18). These methods are not universally applicable to the Magruder check sample program (20) were used as single-
all urea-based fertilizers. They require several reagents, utilize laboratory validation materials, and are listed in Table 1. The
expensive and/or specialty equipment, and are prone to above- validation materials were selected to represent a wide variety
normal operator error due to multistep processes when applied of commercially available urea-based dry and liquid fertilizer
to urea-based fertilizers because of interferences by other products from different sources and manufacturers. The
constituents (19).
materials were used without any further modification, except
As a consequence, no suitable quick and accurate official
that the SCU samples were crushed before use.
method for determining the biuret content in dry and liquid
Nine products were selected for interlaboratory validation: six
commercial fertilizers containing the described urea-
commercially available urea samples, Urea-1, Urea-7, Urea-8,
formaldehyde reaction products exists for regulatory purposes.
An extension of AOAC 2003.14  (19), an LC method for the Urea-9, Urea-10, and Urea-11 from Table 1 (row 9 and rows 17
analyses of urea in select water-soluble and aqueous commercial through 21) plus two commercially available urea-based liquid
urea solutions, is proposed for the determination of biuret in fertilizers, Liquid–5 and Liquid–8 (rows 5 and 8), and one
certain aqueous and dry fertilizer materials. An added advantage urea-based fertilizer sample from the Magruder Check Sample
of the proposed method is the ability to simultaneously determine Program (20) (row 23). They are described in Table 1. All sample
the biuret concentration and the free urea. Excess biuret in urea materials were used without any further modification.
and/or the urea-based fertilizers is detrimental to some plants on
one hand, while on the other, free urea cannot be claimed as slow- Proposed Method (Extension of AOAC 2003.14)
release N for reporting purposes. The proposed method utilizes an
amine chromatography column that was used to separate biuret Scope
from numerous urea-formaldehyde reaction products in these
fluid fertilizers, from urea in commercial urea-based solution This method is applicable to the determination of the biuret
samples, and from urea in sulfur-coated urea (SCU) products. content in both solid and liquid urea-based commercial fertilizer
materials.
Validation Materials
A. Principle
A total of 23 samples, 11 commercial dry urea fertilizer
products, two urea ammonium nitrate products, eight liquid A test portion of homogenous biuret or biuret-containing
urea-based commercial fertilizers, and four SCU samples, from fertilizer sample is diluted with 85% (v/v) acetonitrile and water.

Table  4.  Recovery for validation samples Liquid-5, -8, -9, and -10 spiked with 0.0002 g of biuret
Description Weight, g/100 mL Peak area STDVa g/100 mL Expected, g/100 mL Difference, g/100 mL Recovery, %b

Liquid–8 0.1110 71.9 0.669 0.0002

Spiked 0.1082 126.4 0.717 0.0004 0.0004 0.0002 96.4

Liquid–9 0.1222 12.7 1.189 0.0000

Spiked 0.1151 68.7 2.336 0.0002 0.0002 0.0002 95.5

Liquid–5 0.1134 75.8 0.684 0.0002

Spiked 0.1155 133.6 0.843 0.0004 0.0004 0.0002 99.0

Liquid–10 0.10664 8.2 0.501 0.0000

Spiked 0.1090 65.9 1.364 0.0002 0.0002 0.0002 97.8
a
  STDV = Standard deviation.
b
  Average recovery = 97.2%
 Hojjatie & abrams: journal of aoaC international Vol. 97, no. 3, 2014 715
Figure 1.Example of the standard curve

Table 5. Method precision Biuret Standard Curve

(Average of Seven Analyses)
160.0
Biuret in sample, % 140.0

a 120.0
Fertilizer Wt 1 Wt 2 Wt 3 Wt 4 Average SD RSD, %
100.0 y = 313606x + 1.7604

Peak Area
R² = 0.9999
Liquid–8 0.2045 0.2018 0.1999 0.2019 0.202 0.002 0.945 80.0

60.0
Liquid–9 0.0258 0.0292 0.0323 0.0322 0.030 0.003 10.314 40.0

20.0
Liquid–5 0.2068 0.2095 0.2072 0.2109 0.209 0.002 0.916
0.0
0.0000 0.0001 0.0001 0.0002 0.0002 0.0003 0.0003 0.0004 0.0004 0.0005 0.0005
Liquid–10 0.0204 0.0213 0.0193 0.0228 0.021 0.001 7.111 Concentration of Biuret (g/100ml)

Urea–1 0.684 0.618 0.595 0.576 0.618 0.047 7.615 Figure 1. An example of a standard curve. x = Concentration of
biuret g/100 mL and y = peak area.
Urea–7 — 0.221 0.246 0.242 0.236 0.013 5.682

Urea–8 0.273 0.261 0.314 0.275 0.281 0.023 8.197

A conditioned column will usually last for about one year,
Urea–9 0.636 0.649 0.747 0.769 0.700 0.067 9.639
depending on the number of analyses.
Urea–10 1.033 0.857 0.792 0.751 0.858 0.123 14.344
Urea–11 — 0.731 0.710 0.630 0.690 0.053 7.720 C. Reagents
a
Wt = Weight.
(a) Mobile phase.—85% (v/v) acetonitrile in water. Use LC
grade acetonitrile having 190 nm maximum UV cutoff and LC
Biuret is determined via LC using UV detection at 195 nm. grade water. Caution: Acetonitrile is toxic and flammable.
Biuret is determined by comparison to a standard curve. (b) Biuret > 97% reagent grade.—Cat. No. B54207-
20G (Aldrich Chemical Co.), for use as a reference standard.
B. Apparatus Alternatively, the commercial biuret can be purified by
recrystallization from methanol. The commercial biuret
(a) Liquid chromatograph.—A liquid chromatograph capable standard is placed in methanol and heated. Any insoluble
of isocratic delivery of mobile phase at 2 mL/min at 204 bars material is filtered out, the supernatant solution is allowed to
(3000 psig) and having a UV absorption detector capable of cool, and the needle-shaped crystals are collected by vacuum
stable operation at 195 nm (acetonitrile and water absorption filtration using Whatman filter paper (e.g., 0.45 µm porosity).
cutoff). Operating conditions were flow rate, 1.30 mL/min; The crystallization should yield biuret with a purity near 99.6%
mobile phase temperature, ambient; column temperature, 35°C; (as determined by LC).
detector wavelength, 195 nm; and injection volume, 10 µL. (c) Biuret calibration stock standard.—Accurately weigh
For best precision, a fixed volume sample loop is preferred 0.0048 ± 0.0005 g of biuret [C(b)]. Record the weight to
to syringe injection of samples and standards. If analyzing the nearest 0.1 mg and add to a 100 mL class A volumetric
urea solution for biuret, allocate 14 min for each injection with flask. Add mobile phase to the flask and transfer the flask to
12 min for run time and 2 min for post-run time. For more an ultrasound bath (Cat. No. 98000-346; VWR International,
complex fertilizer solutions, allocate 43 min for each injection Visalia, CA) running at room temperature (approximately
with 23 min for run time and 20 min for post-run time to avoid 25°C). After 4 min, remove the flask from the bath. The flask
overlapping. In this study, a Hewlett Packard HP 1100 Series should be around room temperature (i.e., 25°C). If it is not,
(Agilent Technologies, Santa Clara, CA) liquid chromatograph wait 10–15 min to reach room temperature before adjusting the
was used. The instrument consists of isocratic, binary, and volume to 100 mL. Five calibration standards are prepared by
quaternary pumps (QuartPump); a degasser; an autosampler; a diluting 2, 4, 6, 8, and 10 mL of this stock standard solution
thermostatted column compartment (ColComp); a diode array with the mobile phase to 100 mL in class A volumetric flasks,
detector, a printer, and a computer. resulting in working calibration standards of 0.0001, 0.0002,
(b) LC column.—A 250 × 4.6 mm id amine (NH2) column 0.0003, 0.0004, and 0.0005 g/100 mL of biuret, respectively.
with 5 µm particle size is used for this separation (Phenomenex, All standards were prepared and analyzed within 3–4 days,
Torrance, CA; Part No. 00G-00051-E0). Before use, new Figure 2. Biuret HPLC Chromatogram

columns must be conditioned as described in Section (c).

(c) Column conditioning.—If the LC column is new or has
not been in service for ≥1 week, it must be conditioned before
use as follows:
(1) Using the LC instrument, wash the column for 4 h at
room temperature with LC-grade isopropanol at a flow rate that
will maintain at least 200 bars column backpressure (typically
about 1 mL/min).
(2) Wash the column again for 4 h at room temperature with
100% LC-grade acetonitrile at the flow rate in step 1.
(3) Wash the column a final time with mobile phase
solution [C(a)] at 1.3 mL/min and normal analytical operating
conditions. This wash should continue for at least 2 h or as long
as necessary to obtain a stable baseline. Figure 2. Biuret chromatogram.
716 Hojjatie
Figure & abrams
3. A Representative Chromatogram
: journal of Urea, Biuret, and
of aoaC Urea-forms
international Vol. 97, no. 3, 2014

Figure 3. A representative chromatogram of urea, biuret, and Figure 4. A representative chromatogram of a urea-based liquid
urea forms [Methylenediurea (MDU), monomethylolurea (MMU), fertilizer.
dimethylolurea (DMU), and hexamethylenetetramine (HMT)].

3 days,Figure
otherwise theyillustration
5. Graphical should of beBiuret
keptMethods
in a freezer. A list of
Comparisons

each was injected seven times in over a total of 3 days, and the validation materials is shown in Table 1.
average SD for each sample is listed in Tables 2 and 3.
E. Calculations
(d) Urea.—ACS reagent grade 99–100% (Cat. No. U5128-
100G; Aldrich Chemical Co.) for use as a reference standard. The following is the equation using an external calibration
Dry overnight in a desiccator at room temperature before use, curve for calculation of biuret:
A-b V
and store in a capped glass container. Biuret, % = × × DF
m W

D. Sample Preparation and Biuret Extraction

where A = peak area of biuret in the test solution; b = y-intercept
of the calibration curve; m = slope of the calibration curve;
Liquid fertilizer samples were received in 6 to 12 oz. clear
V = the total volume of the extraction solution; W = test portion
plastic containers, and solid urea samples were received in zip weigh, g; and DF = dilution factor.
lock plastic bags. SCU samples were crushed and thoroughly For example, for sample 24-105-06 (Urea Sample “E”),
mixed before use. All samples were analyzed as received, with A = peak area of biuret in test solution = 111.6, b = y-intercept
of the calibration curve = 1.7604, m = slope of the calibration
no further preparation except as described below. Into a 100 mL curve = 313 606, V = total volume of the extraction solution =
volumetric flask, accurately weigh a test portion (0.0100 to 100 mL, W = weight of the test portion = 0.0291 g, DF = dilution
0.2000 g) dry or liquid fertilizer containing an estimated amount factor = 1, and percentage of biuret in sample 24-105-06 =
[(111.6 – 1.7604)/313 606] × (100/0.0291) × 1 = 1.204.
of free biuret between 1.000 and 4.750 ppm. Dilute to volume
with mobile phase solution from Section C(a). Transfer the Approach for Data Analyses
flasks to an ultrasound bath for 4 min. If the sample is slightly
murky or particles are present in solution, then filter a portion The analytical procedure outlined in Section D was followed
for sample preparation and biuret extraction. Then, 10 μL of
of prepared sample through a 0.45 µm porosity (or finer) filter
extract solution was injected into the LC system equipped with
(Titan 2 HPLC; VWR International) before injecting into the a UV detector at 195 nm, as described in Section B. The biuret
LC column. Sample extract solutions must be analyzed within concentration was determined using the equation provided in
 Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014  717

Table  6.  Urea and liquid fertilizer samples analyzed by three laboratories
Sample ID Lab 1, % Lab 2, % Lab 3, % Mean, % SD s(r)a RSD(r), %a PRSD, %a HorRat

Urea–1 0.66, 0.68 0.62, 0.68, 0.59, 0.58 0.80, 0.80 0.68 0.11 16.72 4.23 3.95
Urea–7 0.16 0.13, 0.22, 0.24, 0.24 0.23 0.20 0.05 25.71 5.07 5.07

Urea–8 0.24, 0.19 0.27, 0.28, 0.31, 0.28, 0.30 0.36 0.28 0.07 24.14 4.84 4.99

Urea–9 0.49, 0.72 0.64, 0.65, 0.75, 0.77 0.72 0.68 0.12 17.28 4.23 4.08

Urea–10 0.97, 0.86 1.03, 0.86, 0.79, 0.75, 0.71 1.11 0.91 0.16 17.64 4.05 4.36

Urea–11 0.15, 0.61 0.73, 0.71, 0.63 0.81 0.61 0.26 43.32 4.30 10.07

Urea–11 0.61 0.73, 0.63 0.81 0.70 0.08 11.57 4.21 2.75

Liquid–5 0.22, 0.20 0.23, 0.23, 0.20 0.27 0.23 0.03 12.86 4.99 2.58

Liquid–8 0.21, 0.24 0.23, 0.22, 0.21 0.25 0.23 0.02 8.05 4.99 1.62
24-032-05 Magruder 0.11, 0.11 0.13, 0.12 N/A 0.12 0.008 7.06 5.50 1.28
a
  SD s(r) = Repeatability standard deviation, RSD(r) = repeatability relative standard deviation, and PRSD = predicted relative standard deviation.

Section E. The mobile phase consisted of a mixture of 85% column is making acceptable separations. Inject 10 µL of each
(v/v) acetonitrile in water. A calibration curve using a minimum biuret standard until two consecutive injections of each give a
of three points was found to be satisfactory once the linearity of peak area within ±2% for the same standard. Inject 10  µL of
the curve is established by an analyst on an LC system. prepared sample and identify the biuret and urea. If the biuret
peak falls within the range of the high and low biuret standards
Validation Parameters then continue with the calculation, otherwise prepare a new
weight-adjusted sample to permit the biuret peak to fall within
Single-Laboratory Validation the standard range.

Linearity.—The linearity of the calibration curve for this

LOD and LOQ
method follows the equation:

y = mx + b The LOD was calculated from the SDs of seven injections

of the lowest biuret standard obtained over 3 separate days
where y = peak area, m = slope of the calibration curve, (three injections on 1 day, and two on each of the other days)
x = biuret concentration in mg/L, and b = y-intercept of the multiplied by 3, or LOD = 3 times SD. From Table 3, the SD
calibration curve. of 0.0324 times 3 results in an LOD of 0.0972 mg biuret/L. The
The method was linear over the region of 1.00 to 4.50 mg/L, LOQ was determined as 10 times the SD, resulting in a value of
with a correlation coefficient ≥ 0.9999. For daily analyses, 0.324 mg biuret/L.
inject a standard solution at least three times to confirm that the Assuming a sample weight of approximately 0.1050 g and
instrument is stable. Compare the chromatogram with the one a final volume of 100 mL, an instrument LOQ of 0.324 mg/L
obtained during the calibration to confirm identity and retention translates to a sample LOQ of approximately 0.309 mg/kg or
times for peaks representing biuret and urea, and to confirm the 0.03% biuret. The method, therefore, is capable of measuring

Table  7.  Biuret method comparisons

Biuret, % Difference between methods
a b c
Urea sample Sample ID TKI (1) TKI (2) Labs (3) 1–2 1–3 2–3

A Apr 09 0.908 1.07 0.93 –0.16 –0.02 0.14

B Jun 09 0.869 0.99 1.20 –0.12 –0.33 –0.21

C Jul 09 1.288 1.36 1.27 –0.07 0.02 0.09

D Dec 09 0.875 1.05 1.10 –0.17 –0.22 –0.05

E Feb 09 1.204 1.28 0.99 –0.08 0.21 0.29

F Oct 09 0.666 0.83 0.81 –0.16 –0.14 0.02

G May 09 1.026 1.11 1.04 –0.08 –0.01 0.07

H Nov 09 1.463 1.55 1.68 –0.09 –0.22 –0.13

Average –0.12 –0.09 0.03
a
  1 = HPLC/Tessenderlo Kerley, Inc. [TKI; current proposed single-laboratory validation (SLV) method].
b
  2 = AOAC Method 976.01 (AAS/by TKI).
c
  3 = Spectrophotometry method (AOAC Method 960.04).
718  Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014

commercially available dry urea samples (Urea–1 and Urea–7

1.80
Comparison of Method for Analysis of Biuret
through Urea–11 from Table 1), with four replicates. The RSDs
for the liquid fertilizer analyses ranged from 0.92 to 10.31%. For
1.60

1.40
the dry urea analyses, the RSDs ranged from 5.68 to 14.34%.
Average Conc. Of Biuret in Sample (%)

1.20 Results are tabulated in Table 5.

1.00
Examples of representative chromatograms for biuret
analyses by LC are shown in Figure 2 (biuret standard), Figure 3
(triuret, biuret, urea, and ureaforms), and Figure 4 (urea-based
0.80

( 1 ) - (HPLC / TKI Method)

0.60 ( 2 ) - AOAC Method 976.01 (AA / by TKI)
( 3 ) - Spectrophotometry (AOAC Method 960.04)
liquid fertilizer).
0.40
CF April 2009 Jun-09 Jul-09 Dec-09 Feb-09 Oct-09 May-09 Nov-09
Sample IDs

Figure  5.  Graphical illustration of biuret method comparisons. Interlaboratory Results

TKI = Tessenderlo Kerley, Inc. (current proposed SLV method).
The standards and validation materials were extracted
biuret well below the critical value of 0.3% recommended by in the Study Director’s laboratory and distributed to two
agronomists. other laboratories from the same stock solutions to maintain
The standard curve was obtained using standard biuret from their uniformity. In addition, interlaboratory studies for six
the Aldrich Chemical Co. [C (b)]. An example of the standard commercial urea samples, two commercial urea-based liquid
curve is shown in Figure 1. fertilizers, and one Magruder check sample were carried out
by these two independent laboratories and the Study Director’s
Instrument Precision laboratory simultaneously. Initially, three practice samples
containing known amounts of biuret were tested by these
Instrument precision was determined based on the repeated laboratories for familiarity before analyzing the blind samples.
injections of biuret standards. The mean peak area, SD, and The comparisons of the results of the three laboratories were
RSD are tabulated in Table 3. For peak areas ranging from 29.50 satisfactory, and no problems were observed due to sample
to 148.40, SD ranged from 0.27 to 1.02, and RSD varied from preparation at the Study Director’s laboratory.
0.30 to 3.34%. The reported values of each duplicate blind sample from
Figure 6. A Representative Chromatogram for Biuret Separation in SCU Samples the collaborative study laboratories were compared for
Recovery reproducibility, linearity, and accuracy. Tabulated data for the
method are shown in Table 6.
Recovery of biuret was determined by spiking slow-release SDs varied from 0.02 to 0.26%, and RSD values ranged from
fertilizer products with biuret and analyzing them according 7.06 to 43.32%, with one value a Grubbs’ outlier (43.32%).
to the method. Four commercially available fertilizer samples,
With this value, the mean of the RSD values was 18.43%.
Liquid–5, Liquid–8, Liquid–9, and Liquid–10 (rows 5, 8, 9, and
When that value was removed, the RSDs ranged from 7.06 to
10 in Table 1), were selected for spiking. The biuret content
25.71%, with the mean of the RSD values equal to 15.67%.
of the test portions was determined to be 0.000237, 0.000224,
0.000037, and 0.000022 g/100 mL, respectively. Test portions Grubbs’ outlier calculation showed the values 24.14 and
were spiked with 0.000188 g/100 mL biuret. Spiked recoveries 25.71% appeared different than the rest, but were not significant
ranged from 95.5 to 99%, with a mean recovery of 97.2%. outliers (P = 0.05). The HorRat (R) values ranged from 1.28 to
According to the results tabulated in Table 4, the recovery of the 5.07%. The HorRat (R) average was 3.41% when the outlier
method is about 97%. was removed.
These results indicate good reproducibility for these low
Method Precision concentration values. The exact concentrations of biuret in
commercial urea samples are not known. Manufacturers
Method precision was determined by analyses of four of urea produce a range of concentrations. The results of
liquid samples (Liquids 5, 8, 9, and 10 from Table 1) and six analyses of these samples were within the expressed ranges.

Table  8.  Statistical values

TKI (1) TKI (2) Labs (3) Mean SD 2SD Mean – 2SD Mean, %
0.908 1.070 0.930 0.969 0.088 0.176 0.793 82

0.869 0.990 1.200 1.020 0.167 0.335 0.685 67

1.288 1.360 1.270 1.306 0.048 0.095 1.211 93

0.875 1.050 1.098 1.008 0.117 0.234 0.773 77

1.204 1.280 0.990 1.158 0.150 0.301 0.857 74

0.666 0.830 0.810 0.769 0.090 0.179 0.589 77

1.026 1.110 1.039 1.058 0.045 0.090 0.968 91

1.463 1.550 1.680 1.564 0.109 0.219 1.346 86
 Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014  719

Table  9.  Biuret in SCU samples Table  10.  Retention times for sulfur, urea, and biuret on
the two columns
LC column
Column retention time, min
Biuret, % Biuret, %
Sample ID Compound Spherex NH2 X-Bridge amide Compound Phenomenex Waters

25-012-04 SCU 0.48 0.49 Sulfur 2.09 2.02

25-012-05 TCS-1 0.71 0.69 Biuret 3.09 3.66
25-012-06 TCS-2 0.83 0.81 Urea 3.71 4.03
25-012-07 XCU 1.01 0.96

turf fertilizer, 39-0-0, S only, S approximately 14% S; and XCU

(a polymer-coated SCU), 43-0-0, S approximately 3.5% S. All
The interlaboratory studies for both urea matrixes and the
samples were crushed and mixed thoroughly before use.
commercial liquid urea-based fertilizers correlated nicely.
The columns used were Waters X Bridge Amide, 3.5  µm,
4.6 × 250 mm column, Part No. 186004870 (Waters Corp.
Comparative Studies
Milford, MA), and Phenomenex column: Spherex 5  µm NH2
(amino propyl), 4.6 × 250 mm column, Part No. 00G-0017-EO,
Two AOAC existing methods are available for biuret analyses
S No. 627458-1 (Phenomenex, Torrance, CA).
in commercial urea samples and urea-based fertilizers. AOAC
SM The results of the analyses of the SCU samples for their
Official Method   960.04 is a spectrophotometric method
biuret contents are shown in Table 9.
based on Cu complexation using alkaline tartrate and CuSO4
Both columns adequately separated the sulfur peak from the
solutions. Other reagents such as NaOH, indicator solution, and
biuret peak for the SCU. However, the Waters column separated
sulfuric acid solution are required for the multistep extraction
and complexation. Standard curves are prepared using biuret the biuret from other peaks, including sulfur, more distinctly.
standard, and the biuret in urea samples or in fertilizers are then The retention times for S, urea, and biuret for each column are
determined spectrophotometrically. shown in Table 10. Figure 6 shows chromatograms representing
AOAC Official Method
SM
976.01 is also based on Cu analyses by both columns.
complexation with subsequent determination by AAS. This
method also needs several reagents, including CuSO4 solution, Conclusions
KOH, KCl, starch, oxalic acid, and indicator solutions.
Comparative studies were performed using the two Urea is the most frequently used source of N fertilizer
aforementioned AOAC methods and the proposed LC method. worldwide. Biuret is a contaminant formed during the production
The results from the studies by the Fertilizer Study Group of urea. Too much biuret is detrimental to plants (1–9). It is
Laboratories (21) on eight commercial urea samples from eight therefore critical to establish a direct and accurate method for
different manufactures by AOAC Method 960.04 were used for the determination of biuret in urea-based fertilizers.
the comparative studies. The LC and AOAC 976.01 methods The proposed LC method is quantitative and capable of
were carried out on the same samples in the Study Director’s determining biuret in liquid urea-based fertilizers (such as urea
laboratory. The results are shown in Table 7. The differences triazone and ureaform), urea solutions, and solid urea-based
among the results of the three methods are shown in Table 7 and
depicted graphically in Figure 5.
When the LC method was paired with AAS method (AOAC
976.01) and the spectrophotometric method (AOAC 960.04),
the P values used was 0.05. The P value was also 0.05 when the
two AOAC methods were compared. Based on the comparisons,
the LC and the AOAC 960.04 method data were statistically
similar (t-test, P = 0.05). Also, the two AOAC methods were
similar. When the AAS method (AOAC 976.01) was compared
to the LC method, it appeared that either the AAS method shows
a high bias or the HPLC method shows a low bias, but in the
absence of a certified material, it is hard to come to a definitive
conclusion. The statistical values for the three methods are
shown in Table 8.

SCU Sample Analyses: Interference Testing

SCU was investigated as a possible interferent in the

determination of biuret. To test this, four SCU materials were
analyzed for biuret by the proposed method using two different
columns. The materials and their respective S concentration are
as follows: SCU (S approximately 2%), ground; TCS-1 (a turf Figure  6.  Representative chromatograms for biuret separation in
fertilizer), 39-0-0, S + Wax, S approximately 14% S; TCS-2 (a SCU samples.
720  Hojjatie & Abrams: Journal of AOAC International Vol. 97, No. 3, 2014

fertilizers (such as different SCU samples). The performance   (3) Mikkelsen, R. (1990) Nutr. Cycl. Agroecosys. 26, 311–318
characteristics of the method were established in terms of   (4) Hasan, M.M. (1957) Biuret Phytotoxicity, UCLA-Horticultural
repeatability. The LC results were similar to those of the Science Publication, Los Angeles, CA
AOAC  960.04 and AOAC 976.01 methods; however, those   (5) Kilmer, V.J., & Engelstad, O.P. (1973) National Fertilizer
methods are more tedious and generate hazardous Cu waste Development Center Bulletin Y-57, Muscle Shoals, AL
solution.   (6) Webster, G.C., Berner, R.A., & Gansa, A.N. (1957) Plant
In addition, the proposed method has already been established Physiol. 32, 60–61. http://dx.doi.org/10.1104/pp.32.1.60
  (7) Achor, D.S., & Albrigo, L.G. (2005) J. Am. Soc. Hort. Sci. 130,
as Official MethodSM 2003.14 for the determination of unreacted
667–673
free urea in urea-based fertilizers. Free urea is not considered a
  (8) Singh, D., Singh, M., & Sandhu, H.S. (1979) Indian J. Agric.
source of slow release N, and cannot be claimed as such. An
Sci. 49, 641–648
advantage of the proposed method is that it can be used to   (9) Abdel-Hadi, A.H., Khadr, M.S., & Bakr, M.N. (1983) Z.
simultaneously determine the amount of biuret and free urea in Pflaz Bodnek 146, 379–384. http://dx.doi.org/10.1002/
such fertilizers for regulatory reporting purposes. jpln.19831460313
Furthermore, since fertilizers manufactured from urea (10) Walker, J.F. (1964) Formaldehyde, 3rd Ed., ACS Monograph
contain some lower chain ureaforms such as monomethylolurea, Series, Reinhold Publishing Co., New York, NY, Chapter 14,
dimethylolurea, methylenediurea, and triuret, the proposed pp 359–414
method could also be extended to determine these moieties (11) Clark, K.G., Yee, J.Y., & Love, K.S. (1948) Ind. Eng. Chem.
individually or simultaneously. This is proposed as future work. 40, 1178–1183
(12) Clark, K.G., Yee, J.Y., Love, K.S., & Boyd, T.A. (1951)
Acknowledgments Ind. Eng. Chem. 43, 871–875. http://dx.doi.org/10.1021/
ie50463a004
We give special thanks to Hugh Rodrigues, Thornton (13) Hauck, R.D., & Koshino, M. (1971) Fertilizer Technology
and Use, 2nd Ed., R.A. Olson (Ed.), Soil Science Society of
Laboratory, Tampa, FL, and John Hartshorn, Morral Chemical
America, Madison, WI, pp 455–494. http://dx.doi.org/10.1021/
Co., Morral, OH, for collaborative support of this study. We
ie50496a031
are also indebted to James Bartos of the Office of Indiana State
(14)  Official Methods of Analysis (2000) 17th Ed., AOAC
Chemist and Nancy Thiex for many helpful suggestions. We
INTERNATIONAL, Gaithersburg, MD, Method 960.04
are also thankful to Harold Falls of CF Industries for providing (15) Ellis, G.C., & Formaini, R.L. (1955) J. Agric. Food Chem. 3,
some of the commercial urea test samples and the test results 615–618
for biuret analyses by AOAC Official MethodSM  960.04. We (16) Xiaoyan, Z., Congge, X., Qlaoying, Z., & Zhonghang, W.
also thank Hugh Rodrigues (Thornton Laboratory) and Fred (1992) Nutrient Cycling in Agroecosystems 33, 93–96
Carney (Agrium Advanced Technology) for providing the SCU (17) Farlin, S.D., Schelling, G.T., & Garrigus, U.S. (1967) J. Anim.
samples. Sci. 26, 1205–1209. http://dx.doi.org/10.1021/jf60053a008
(18) Geurts, J.J., Van Stelle, J.E., & Brinkman, E.G. (1968) Anal.
References Chim. Acta 41, 113–120
(19) Hojjatie, M.M., & Abrams, D.E. (2004) J. AOAC Int. 87,
  (1) Redemann, C.E., Riesenfeld, F.C., & Viola, F.S. (1958) 346–352. http://dx.doi.org/10.1016/S0003-2670(01)80367-6
Ind. Eng. Chem. 50, 633–636. http://dx.doi.org/10.1021/ (20) Magruder Fertilizer Check Sample Program, Sample 2012-05A.
ie50580a035 magruderchecksample.org
  (2) Shen, R.C.J. (1959) J. Agric. Food Chem. 7, 762–763. http:// (21) Fertilizer Study Group Laboratory Test Results (2012), provided
dx.doi.org/10.1021/jf60105a002 by Harold Falls, CF Industries, Plant City, FL
Copyright of Journal of AOAC International is the property of AOAC International and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.