50.1.25 (b) Column.

—Any C18 column (monomeric or polymeric) con-

AOAC Official Method 999.15 taining 5 µm spherical particle silica with ≥10% carbon loading, pre-
Vitamin K in Milk and Infant Formulas ceded by a guard column of similar packing.
Liquid Chromatographic Method (c) Post-column reductor.—20 × 4 mm stainless steel
First Action 1999 post-column assembly placed between analytical column and fluo-
rescence detector (Part No. WAT084550, Waters, or equivalent).
(Applicable to the determination of total vitamin K1 [phylloquinone] Alternative devices, such as an empty (30–50 mm) LC column can
in infant formula and milk [fluid, ready-to-feed, and powdered] con- be used.
taining >1 µg vitamin K1/100 g solids.) (d) Spectrophotometer.—UV/Vis spectrophotometer, digital
readout to 0.001 A.
See Table 999.15 for the results of the interlaboratory study sup- (e) Waterbath.—37 ± 1°C.
porting the acceptance of the method. (f) Centrifuge.—Fixed or swing-arm rotor.
(g) Mechanical shaker.—Orbital or wrist-action.
Caution: Avoid skin contact with zinc dust.
(h) Vortex mixer.
A. Principle (i) pH meter.—Accurate to 0.01 pH, with calibration buffers.
(j) Test tubes.—200 × 24 mm with ground-glass stoppers.
Following enzymatic digestion of fat and precipitation of fatty ac-
(k) Vials.—Amber-tinted glass vials (1.8 and 8.0 mL), with Tef-
ids, vitamin K is extracted with hexane. Vitamin K is separated by
lon-sealed caps.
HPLC with post-column reduction and quantitated by fluorescence
(l) Volumetric flasks.—Low-actinic volumetric flasks, 100 mL.
detection with an external standard.
Phylloquinone geometric isomers are quantitated either as a sin-
C. Reagents
gle unresolved peak with a C18, or may be estimated selectively with a
C30 column (k′trans < k′cis), under the reversed-phase conditions em- Use AR grade reagents unless otherwise stated.
ployed. (a) Water.—Purified to >18 MΩ resistivity.
(Note: Menaquinone-4 [MK-4] will also be present in milk and in- (b) Nitrogen.—Cylinder N (low O content).
fant formulas incorporating milk fat, while 2′,3′-dihydrophylloquinone (c) Lipase.—From Candida rugosa, ca 1000 units/mg (Type VII,
may be present at appreciable levels in infant formulas containing hy- Sigma L-1754). Other sources of lipase can be used from Pseudomo-
drogenated soya or canola oils. If required, both may be estimated nas and Rhizopus species with the correct activity of enzyme.
against appropriate standards. The order of elution is MK-4 < K1 < (d) Monobasic potassium phosphate (KH2PO4).
2′,3′-dihydroK1 [relative retention times ca 0.6, 1.0, and 1.2, respec- (e) Potassium hydroxide solution.—40% (w/v). Dissolve 40 g
tively].) KOH, (j), in water and dilute to 100 mL.
(f) Phosphate buffer.—0.8M. Dissolve 54.0 g KH2PO4, (d), in
B. Apparatus 350 mL water. Adjust pH to 7.9–8.0 with 40% KOH solution and di-
(a) HPLC system.—Manual or automated isocratic system incor- lute to 500 mL.
porating a fluorescence detector (λex 243 nm, λem 430 nm) exhibiting (g) Ethanol.
a Raman spectra signal:noise specification of >200:1 (water). Since (h) Methanol.—LC grade.
UV detection is insensitive to vitamin K at the working concentra- (i) Reagent alcohol.—Mix ethanol, (g), with methanol, (h),
tion levels, configure with single-mode fluorescence detection only (95 + 5, v/v).
to minimize extra-column band-broadening. For multi-detector (j) Potassium hydroxide.
LCs, configure with the fluorescence detector in the first position. (k) Zinc chloride.
Ensure the absence of leaks. (l) Sodium acetate.—Anhydrous.

Table 999.15 Interlaboratory study results for vitamin K1 collaborative data

Samplea
Parameter 1 2 3 4 5 6 7 8
No. of labsb 31 28 33 33 33 33 32 33
No. of outliers 1 5 0 0 0 0 1 0
Mean, µg/100 g 0.49 6.63 118.07 32.24 78.69 49.64 90.94 94.62
sr, µg/100 g 0.04 0.21 5.00 1.54 2.04 2.54 4.04 5.38
RSDr, % 9.03 3.23 4.24 4.77 2.59 5.11 4.44 5.68
2.8 sr, µg/100 g 0.12 0.60 14.01 4.31 5.71 7.11 11.32 15.05
sR, µg/100 g 0.05 0.39 6.50 2.14 3.40 3.80 4.14 6.41
RSDR, % 10.94 5.81 5.50 6.63 4.33 7.66 4.56 6.78
2.8 sR, µg/100 g 0.15 1.08 18.19 5.98 9.53 10.65 11.60 17.95
a
Samples 1 and 2 = unfortified, whole liquid UHT milk and goat milk powder, respectively; 3 = milk-based infant formula, oil-filled; 4 and 7 = whey-based infant formula,
partially oil-filled; 5 = soy-based infant formula, oil-filled; 6 = whey-based infant formula, oil-filled; 8 = NIST SRM 1846.

© 2000 AOAC INTERNATIONAL

 (m) Acetic acid.—Glacial. to ambient temperature by standing in water. Add 10 mL reagent al-
(n) Potassium carbonate.—Anhydrous. cohol and mix. Add 1.0 g K2CO3 and mix.
(o) Zinc powder.—<60 µm, e.g., Merck 108774 and BDH (2) Extraction.—(Note: Take precautions to prevent concentra-
102964L. Use freshly opened bottle. tion of extracts through uncontrolled solvent evaporation.) Add
(p) Hexane.—HPLC grade. 30.0 mL hexane, stopper, and shake vigorously for ≥10 min using
(q) Dichloromethane.—HPLC grade. mechanical shaker. Let stand in the dark, preferably refrigerated, un-
(r) Isopropanol.—HPLC grade. til layers separate. If this is slow (>10 min) or incomplete, centrifuge
(s) Mobile phase.—To dichloromethane, (q),-methanol, (h), at ca 1000 rpm (ca 200 g) for 10 min (decant a portion of supernatant
(100 + 900, v/v), add 5 mL methanol containing ZnCl2 (1.37 g; k), into smaller centrifuge tube if necessary). Transfer 1.00 mL (forti-
anhydrous sodium acetate (0.41 g; l), and glacial acetic acid (0.30 g; fied products) or 5.00 mL (nonfortified products) of upper hexane
m). Filter through 0.45 µm filter and degas. phase into vial. Evaporate under N flow to dryness. (Note: larger ex-
(t) Zinc post-column reductor.—Dry-pack the 20 × 4 mm assem- tract volumes may be required, with the option of rotary evaporation,
dependent on the sensitivity of the fluorescence detector.) These
bly B(c), with Zn powder, (o). Minimize voids by sequentially add-
dried extracts may be held for 3–5 days in the dark at <0°C under N
ing small amounts with frequent gentle tapping. With longer
prior to LC analysis. Re-dissolve residue in 1.00 mL methanol. Seal
columns, use extra care. Re-prepare reductor column before each
vial and Vortex mix. If using autosampler, use a low volume vial (or
analytical schedule. Reduction efficiency of Zn powder should be
insert) to minimize evaporation into the headspace.
monitored over time and replaced every 1–2 years as required.
(b) HPLC determination.—(1) Set-up.—Prior to connection of
(u) Vitamin K1 (phylloquinone).—USP grade. Conduct all opera-
post-column Zn reductor assembly, establish stable operating LC
tions must be conducted under low incandescent light conditions
conditions over ca 30 min through equilibration of column with mo-
and in low-actinic volumetric flasks. (1) Stock solution.—ca
1.0 mg/mL. Dissolve ca 100 mg, weighed accurately, in 100.0 mL bile phase (ca 1 mL/min) and set fluorescence detector to λex 243 nm
isopropanol (r) with warming at ca 30°C and standing. Store under and λem 430 nm (gain and sensitivity will be detector dependent). In-
N, (b), at –10°C for up to 6 months. (2) Intermediate I.—ca stall Zn reductor assembly between analytical column and detector
and equilibrate system for a further 30–60 min, ensuring the absence
50 µg/mL. Dilute 5.0 mL stock with methanol to 100.0 mL and store
of leaks (anhydrous LC eluent conditions are required to minimize
under N at –10°C for up to 1 month. (3) Intermediate II.—ca
loss of reducing potential of Zn). Do not recycle mobile phase. Inject
2.5 µg/mL. Dilute 5.0 mL Intermediate I with methanol to
a standard (20–50 µL) and set flow rate in order to elute
100.0 mL. Prepare daily. (4) Working standards.—ca 6.25, 12.50,
phylloquinone between 8–15 min, at typically 1.0–1.5 mL/min (re-
18.75, 25.00, and 31.25 ng/mL. Dilute 0.25, 0.50, 0.75, 1.00, and
tention may also be manipulated by modification of dichloro-
1.25 mL, respectively, Intermediate II with methanol, (h), to
methane content in eluent between 8–12%). Allow total run time of
100.0 mL. Prepare daily and calculate accurate concentrations as
ca 5 min beyond elution of K1.
follows: Accurate concentrations for working standards are calcu-
(2) Calibration and analyses.—Inject methanol (standard blank)
lated after estimation of phylloquinone purity from absorbance
and confirm absence of chromatographic activity at retention time
(248 nm) of a solution prepared by evaporating 5.00 mL Intermedi-
for phylloquinone. Inject lowest level working standard and ensure
ate I under N flow and re-dissolving in 25.0 mL hexane (a1% = 419):
repeatable response, with signal:noise ≥10. Inject each working
standard and ensure linear response. Sequentially inject reagent
W 5 5
Theoretical A248 = × × × 419 × 100 blank and test extracts and include a working standard after every
100 100 25 4–6 test solutions to monitor system stability. When analytical
schedule is complete, flush system with methanol. Store analytical
where W = weight of phylloquinone in stock standard (g) column in methanol when not in use. Dismantle and empty
Purity factor (PF) = measured A248/theoretical A248. post-column assembly, soak in 5M HCl to clean frits, and rinse thor-
oughly with water and methanol. Dry thoroughly before repacking
W 5 106 with Zn.
K1 in Intermediate II (µg/mL) = PF × × ×5 ×
100 100 100 (c) Calculation.—Construct a 5-level calibration using either
peak area or height and calculate slope (S) by linear regression with
Accurate vitamin K1 concentrations in the 5 working standards forced zero. Calculate vitamin K1 (cis- and trans-) content in test
(ng/mL) are then calculated by multiplying PF by 2.5, 5.0, 7.5, 10.0, samples as follows:
and 12.5, respectively.
A′ 30 100 1
Vitamin K1 (µg/100 g) = × × ×
D. Procedure S V Wt 1000
Perform all steps of assay under incandescent lighting in the ab-
sence of direct sunlight. where A′ = peak area (or height) of phylloquinone in test extract;
(a) Preparation of test solution.—(1) Digestion.—Weigh 1.0 g Wt = weight of test portion (g); V = volume, 1.0 mL (fortified) or
powder or 10.0 g liquid into test tube. Include both a reagent blank 5.0 mL (nonfortified); S = slope of calibration graph.
and a quality control test portion in each analytical schedule. Dis- Calibration and calculations may be achieved through data pro-
solve powder in ca 15 mL warm water (<40°C) with Vortex mixer cessing within the instrument or off-line.
(for liquids, add 5 mL warm water). Add 5.0 mL phosphate buffer to Reference: J. AOAC Int. 83, 121(2000).
each solution and mix. Add 1.0 g lipase powder and Vortex mix.
Stopper securely and shake until well dispersed (30–60 s). Incubate
at 37 ± 2°C for 2 h. Shake each tube for 15 s at 20 min intervals. Cool

© 2000 AOAC INTERNATIONAL