Method Development
and Validation for Simultaneous Determination of 44Ca, 34S,
28Si and 18 Other Trace Elements in Pharmaceutical Packaging
Materials’ Extractable Solutions by Inductively Coupled Plasma-Mass
Spectrometry (ICP-MS)
Dan Xie*, WeiChun Yang, Qin Lu
R&D Center, Baxter
Healthcare (Suzhou) Co., Ltd., Suzhou, China
*Corresponding Author E-mail: dan_xie@baxter.com
ABSTRACT:
ICH (International Council for
Harmonisation of Technical Requirements for Pharmaceuticals for Human Use)
already issued and implemented Q3D guideline in elemental impurities in final
drug products. Consequently, it will be essential to monitor trace elements
from packaging material to ensure the final drug product compliance. This study
successfully developed a new method for simultaneous identification and
quantification of 44Ca, 34S, 28Si and 18 trace
elements (27Al, 51V, 52Cr, 55Mn, 56Fe,
58Ni, 59Co, 63Cu, 66Zn, 75As,
78Se, 95Mo, 111Cd, 118Sn, 121Sb,
137Ba, 201Hg and 208Pb) in pharmaceutical
packaging materials’ extractable solutions by using ICP-MS in one single method
without auxiliary. The method development focused on elemental mass selection,
optimization of ICP-MS operational parameters and the sample/standard solutions
preparation. Furthermore, the new developed analytical method (accuracy and precision,
standard and sample linearity, matrix specificity and robustness of the method)
was successfully validated by following US and European compendia criteria. The
success of the analytical method development and validation illustrates that
the trace elements analysis in pharmaceutical industry becomes feasible per the
single ICP-MS method. The analysis of trace elements via this new developed
ICP-MS method can provide the worthy information for risk assessment of
packaging system and final drug products with relatively low operational cost.
KEYWORDS: Trace Elements,
Extractable Solutions, ICP-MS, Method Development and Validation.
INTRODUCTION:
The drug product’s packaging
systems, manufacturing components and/or drug administration devices may be
constructed from plastic, elastomeric or glass materials. These materials are
one potential source of elemental impurities. Trace elements presented in such
materials may leach (or be extracted) from these materials while they are in
contact with a parenteral drug product and may accumulate in the drug product,
which could adversely affect the quality, stability, potency and/or safety of
the drug product. It is necessary and appropriate that plastic, elastomeric and
glass materials, components and systems are tested to establish their
extractable trace element and metal profiles1.
Guideline ICH Q3D
(International Council for Harmonisation of Technical Requirements for
Pharmaceuticals for Human Use, Q3D) defines elemental impurities control in
drug and is also the basis of risk management of drug elemental impurities. In
this study, the elements in ICH Q3D guideline are monitored in drug and
pharmaceutical packaging materials’ extractable solutions2. Calcium
(Ca), Silicon (Si) and Sulfur (S) in extractable solutions that are nontoxic or
low toxic elements are indirectly related the safety effects from packaging
system. Calcium is targeted as it may cause the formation of particulate
matter, and Si and S are targeted as indicators for certain organic extractables.
The identification and quantification of Si and S can facilitate the
toxicological assessment of organosulfur and organosilicon extractables1.
Therefore, these trace elements mostly extracted from plastic, elastomer and
glass material components should be identified and quantified by thresholds1.
That is because the toxicity of these elements per ICH Q3D and the inherent
characteristics of all packaging materials, as well as the relevant risk
assessment, are needed2,3.
In theory, Inductively Coupled
Plasma-Mass Spectrometry (ICP-MS) is capable to quantify a variety of trace
elements in aqueous samples. But the identification and quantification of Ca, S
and Si by ICP-MS are considered as a big challenge, because of numerous masses,
ions and matrix interference on ionization and detection. Regarding analysis of
elements Ca and Si in aqueous solutions, the current analytical technologies
focus on the application of Inductively Coupled Plasma-Optical Emission
Spectrometer (ICP-OES)4-5. But ICP-OES with relatively high
detection limit is not the best choice for trace elements analysis in the risk
assessment of extractable elements6-7. Compared with ICP-OES, ICP-MS
is a more advanced technique with lower detection limit and higher accuracy and
precision8. The identification and quantification of Ca, Si and S by
ICP-MS in many research fields followed two strategies9-11.
First, ICP-MS was used to
quantify Ca, S and Si accurately by using auxiliaries10,12,13. The
purpose of ICP-MS combined with auxiliaries is to minimize interference caused
by the inherent character of plasma. The application of high-resolution ICP-MS
in drug substance analysis resulted in high performance to identify element S12.
Trace silicon in biological and clinical samples could be determined by
ICP-high resolution isotope dilution MS10. The trace
level Ca in food was identified by ICP-MS equipped with a dynamic reaction cell
(DRC)13. Although the analytical method development of trace
elements analysis with auxiliaries generated good results, it led to relatively
high experimental cost and instrument investment. Consequently, these advanced
ICP-MS methods cannot be widely applied in the industry.
On the other hand, when the
interference cannot be eliminated, the ‘ordinary’ ICP-MS without extra
auxiliaries could only semi-quantify Ca14. The semi-quantification
method for the trace elements analysis with relatively low experiment cost used
in the industry cannot completely meet the requirements to assess the elemental
risks caused by packaging materials. In particular, it is a big challenge to
reach nanogram per milliliter level for elements S and Si identification and
quantification.
Under the current situation,
one simple method for Ca, S and Si analysis and other 18 extracted trace
elements (Al, V, Cr, Mn, Fe, Ni, Co, Cu, Zn, As, Se, Mo, Cd, Sn, Sb, Ba, Hg and
Pb) in extractable solutions from pharmaceutical packaging material by
using ICP-MS without auxiliaries is eagerly needed in the pharmaceutical
industry. The purpose of this study is to develop and validate a method for
simultaneous identification and quantification of 21 trace elements by ICP-MS
in a single method without any auxiliary.
MATERIAL AND METHODS:
Chemicals and reagents:
All chemicals were analytical
grade or above. Chemicals and stock standard solutions stored in polypropylene
material containers was used to avoid metal interference from containers.
Concentrated HCl (TraceMetal, Thermo Fisher Scientific, US) were used for rinse
(10%HCl v/v) and rinse port (10%HCl v/v) solutions preparation. Rinse solution
(10%HCl v/v) was used as the matrix of standard solution and sample
preparations. It was also used for Blank rinse, Probe rinse, Loop wash, and
Carrier tube feed when using Integrated Sample Introduction System (ISIS).
Internal standard (ISTD) contained 0.300µg/mL Y, In, and U (customized ISTD
solution in Table 1) in 10% HCl v/v.
Stock standard solutions shown
in Table 1 were commercially customized/purchased and traceable to National
Institute of Standards and Technology (NIST) reference standards. The stock
standard solutions were used for the preparation of calibration standards in
different concentration levels for individual element based on matrix 10% HCl
v/v. The calibration standard solutions preparation and concentration of each
element are shown in Table 2A and Table 2B.
Table 1. Reference Stock
Standards
|
Reference Standard Name |
Supplier |
Concentration (mg/L) |
Elements |
Standard Matrix |
|
Customized-Std 1 |
Inorganic Ventures |
10.0 |
Cu, Mn, Mo |
10% v/v HCl/tr. HF (trace amount of HF) |
|
5.0 |
Sb, Cr, V |
|||
|
2.5 |
As, Cd, Co, Pb, |
|||
|
Customized-Std 2 |
Inorganic Ventures |
25 |
Fe |
5% v/v HNO3 |
|
15 |
Ba, Ni, Se |
|||
|
Customized-Std 3 |
Inorganic Ventures |
10 |
Sn, Zn |
10% v/v HCl |
|
5.0 |
Al |
|||
|
Customized-Std 4 |
Inorganic Ventures |
1 |
Hg |
10% v/v HCl/tr. HF |
|
15 |
Ca |
|||
|
100 |
S, Si |
|||
|
Single Ca standard |
CPA Chem |
10 mg/L |
Ca |
2% v/v HNO3 |
|
Customized-Std QC |
Inorganic Ventures |
2.5 |
Al, Sb, As, Ba, Cd, Cr, Co, Cu, Fe, Pb, Mn, Mo, Ni, Se, V, Zn, Sn |
10% v/v HCl/tr. HF |
|
0.5 |
Hg |
|||
|
5 |
Ca |
|||
|
25 |
S, Si |
|||
|
Customized Internal Standard (ISTD) |
Inorganic Ventures |
10 µg/mL |
Y, In, U |
5% v/v HNO3 |
|
Customized EI-SPK-1 |
Inorganic Ventures |
50 µg/mL |
Cr, Cu, Mn, Mo, V |
5% v/v HNO3 / tr. HF |
|
25 µg/mL |
Sb, As, Cd, Co, Pb |
|||
|
Customized EI-SPK-2 |
Inorganic Ventures |
50 µg/mL |
Fe, Ni, Ba, Se |
5% v/v HNO3 |
|
Customized EI-SPK-3 |
Inorganic Ventures |
100 µg/mL |
Sn, Zn |
5% v/v HNO3 / tr. HF |
|
50 µg/mL |
Al |
|||
|
Single Ca Spk standard |
CPA Chem |
1000 mg/L |
Ca |
2% v/v HNO3 |
|
Single Si Spk standard |
o2si |
1000 mg/L |
Si |
H2O |
|
Single S Spk standard |
o2si |
1000 mg/L |
S |
H2O |
|
Single 10mg/L Hg standard |
Agilent |
10 µg/mL |
Hg |
5% v/v HNO3 |
Table 2A. Standard
Solution Preparation (unit: mL)
|
Standard solution ID |
2Masshunter Software Setup ID |
1Rinse solution, mL |
Std 1, mL |
Std 2, mL |
Std 3, mL |
Intermediate SolutionmL |
Std 4, mL |
Single Ca std. |
Std QC |
Final Volume with 1Rinse solution, mL |
|
|
20 elements (S excl.) |
S element |
||||||||||
|
Intermediate solution |
N/A |
~25 |
1.000 |
0.500 |
2.500 |
0 |
0 |
0 |
0 |
50 |
|
|
WS0 (Blank) |
Level 1 |
Level 6 |
~25 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
50 |
|
WS1 (Calibration Standard1) |
Level 2 |
Level 7 |
~25 |
0 |
0 |
0 |
1.000 |
0.050 |
0 |
0 |
50 |
|
WS2 (Calibration Standard 2) |
Level 3 |
Level 8 |
~25 |
0.200 |
0.050 |
0.100 |
0 |
0.100 |
0 |
0 |
50 |
|
WS3 (Calibration Standard 3) |
Level 4 |
Level 9 |
~25 |
0.400 |
0.100 |
0.200 |
0 |
0.300 |
0 |
0 |
50 |
|
WS4 (Calibration Standard 4) |
Level 5 |
Level 10 |
~25 |
1.000 |
0.500 |
1.000 |
0 |
0.800 |
0.800 |
0 |
50 |
|
QC (Quality Check) |
QC |
~25 |
0 |
0 |
0 |
0 |
0.500 |
0 |
0.500 |
50 |
|
1 Rinse solution indicated 10%
HCl v/v. 2 Since 20 elements and S in one single sample are
analyzed in two sublists using ICP-MS Masshunter software, the level setup of
WS solutions were different.
Table 2B. Concentration of Standard Solution Prepared (unit: ng/mL)
|
Concentration |
Intermediate solution |
WS0 |
WS1 |
WS2 |
WS3 |
WS4 |
QC |
|
Aluminum (Al) |
250 |
0 |
5 |
10 |
20 |
100 |
25 |
|
Silicon (Si) |
|
0 |
100 |
200 |
600 |
1600 |
1250 |
|
Sulfur (S) |
|
0 |
100 |
200 |
600 |
1600 |
1250 |
|
Calcium (Ca) |
|
0 |
15 |
30 |
90 |
400 |
200 |
|
Vanadium (V) |
100 |
0 |
2 |
20 |
40 |
100 |
25 |
|
Chromium (Cr) |
100 |
0 |
2 |
20 |
40 |
100 |
25 |
|
Manganese (Mn) |
200 |
0 |
4 |
40 |
80 |
200 |
25 |
|
Iron (Fe) |
250 |
0 |
5 |
25 |
50 |
250 |
25 |
|
Nickel (Ni) |
150 |
0 |
3 |
15 |
30 |
150 |
25 |
|
Cobalt (Co) |
50 |
0 |
1 |
10 |
20 |
50 |
25 |
|
Copper (Cu) |
200 |
0 |
4 |
40 |
80 |
200 |
25 |
|
Zinc (Zn) |
500 |
0 |
10 |
20 |
40 |
200 |
25 |
|
Arsenic (As) |
50 |
0 |
1 |
10 |
20 |
50 |
25 |
|
Selenium (Se) |
150 |
0 |
3 |
15 |
30 |
150 |
25 |
|
Molybdenum (Mo) |
200 |
0 |
4 |
40 |
80 |
200 |
25 |
|
Cadmium (Cd) |
50 |
0 |
1 |
10 |
20 |
50 |
25 |
|
Tin (Sn) |
500 |
0 |
10 |
20 |
40 |
200 |
25 |
|
Antimony (Sb) |
100 |
0 |
2 |
20 |
40 |
100 |
25 |
|
Barium (Ba) |
150 |
0 |
3 |
15 |
30 |
150 |
25 |
|
Mercury (Hg) |
0 |
1 |
2 |
6 |
16 |
15 |
|
|
Lead (Pb) |
50 |
0 |
1 |
10 |
20 |
50 |
25 |
Triplicate of each calibration
standard solutions, which contained four different concentrations of standard
solutions and one blank with acidic background solution for each element, were
prepared and analyzed. Quality Check (QC) solution shown in Table 2A were
inserted in the analytical sequence, after approximately every 15 samples
analysis to check the system suitability of the method during the analysis. The
results of standard solution analysis were used to define the methodological
system suitability criteria and standard linearity calculation.
One liter buffer contained
0.825g Potassium Chloride (KCl, Merck, Germany) was prepared and adjusted to pH
2 ± 0.1 using 5N HCl (Merck, Germany). The buffer solution was stored at
ambient temperature and used for method development and validation.
Instrumentation:
This study was performed using
Agilent ICP-MS 7700X (Agilent Technologies, USA) configured with Automatic
sampler (Agilent ASX-500 series or equivalent), High Matrix Introduction (HMI),
ISIS, Octopole Reaction System (ORS) with 100% helium (He) gas in the mode of
discrete sampling (ISIS-DS) analysis, three channel peristaltic pump for
carrier solution (10% HCl v/v), ISTD and drain, Chiller.
The instrument operational
parameters, ICP-MS rinse parameters and element detection parameters were
optimized and stated in Table 3A and Table 3B, respectively.
Table 3A. Optimized Instrument
Operation Parameters
|
Plasma Parameters |
|
|
RF Power |
1450 W |
|
RF Matching |
1.50 V |
|
Sampling Depth |
10.0 mm |
|
Spray Chamber Temp |
2°C |
|
Carrier Gas |
0.65 L/min |
|
Option Gas |
0.0 % |
|
Nebulizer Pump |
0.10 rps |
|
Dilution Gas |
0.60 L/min |
|
Sampler/skimmer cones |
Nickel |
|
Cell Parameters |
|
|
He Flow |
4.5 mL/min |
|
H2 Flow |
0 mL/min |
|
3rd Gas Flow |
0 % |
|
Octopole Bias |
-18.0 V |
|
Octopole RF |
190 V |
|
Meters |
|
|
IF/BK Press |
1.86E+0 Pa |
|
Analyzer Press |
1.97E-5 Pa |
|
Plasma Frequency |
25.00 MHz |
|
Argon Gas Tank Press |
5.91E+2 kPa |
Table 3B. Detection Parameters
for 21 Elements
|
Mass |
Element Name |
Integration Time /Mass [sec] |
Analyte/ Internal Standard (ISTD) |
|
27 |
Al |
1.29 |
Analyte |
|
28 |
Si |
0.99 |
Analyte |
|
34 |
S |
9 |
Analyte |
|
44 |
Ca |
6 |
Analyte |
|
51 |
V |
0.3 |
Analyte |
|
52 |
Cr |
0.51 |
Analyte |
|
55 |
Mn |
0.21 |
Analyte |
|
56 |
Fe |
0.21 |
Analyte |
|
58 |
Ni |
0.3 |
Analyte |
|
59 |
Co |
0.3 |
Analyte |
|
63 |
Cu |
0.21 |
Analyte |
|
66 |
Zn |
0.21 |
Analyte |
|
75 |
As |
1.29 |
Analyte |
|
78 |
Se |
1.29 |
Analyte |
|
89 |
Y |
0.3 |
ISTD |
|
95 |
Mo |
0.3 |
Analyte |
|
111 |
Cd |
1.29 |
Analyte |
|
115 |
In |
0.3 |
ISTD |
|
118 |
Sn |
0.09 |
Analyte |
|
121 |
Sb |
0.21 |
Analyte |
|
137 |
Ba |
0.3 |
Analyte |
|
201 |
Hg |
3 |
Analyte |
|
206 |
Pb |
1.11 |
Analyte |
|
207 |
Pb |
1.11 |
Analyte |
|
208 |
Pb |
1.11 |
Analyte |
|
238 |
U |
0.3 |
ISTD |
|
Acquisition mode: |
Spectrum |
|
|
|
Peak pattern: |
3 |
|
|
|
Replicate: |
3 |
|
|
|
Sweep, replicate |
100 |
|
|
|
Correction equation |
208Pb=1× 206Pb + 1× 207Pb + 1× 208Pb |
|
|
Method development procedure:
Elemental mass selection for
21 elements and optimization of instrumental parameters:
Due to the identifying
difficulty from atomic interference of ICP-MS, this study started from elements
Ca, S and Si identification and quantification 1. Element Ca was
identified by seeking the optimized instrument parameters and the most suitable
detective mass among masses 40, 42 and 44. 40Ar+ from
Argon gas interfered with 40Ca identification, leading to poor
concentration recovery of 40Ca 15. Compared with 40Ca
recovery, 42Ca and 44Ca had better concentration recovery
in four calibration standard solutions. Moreover, the acidic background
solutions 10%HCl (v/v) impacted the identification of 44Ca less than
42Ca. Therefore, regarding element Ca analysis, the research
concentrated on mass 44.
Element S was initially
analyzed masses 32, 33 and 34. As per atomic interference of ICP-MS, NO3-
existed in background solutions had strong interference on S and Si 15.
Therefore, HNO3 was not used as the background solution to eliminate
the matrix interference on element S. In addition, 32O2+
and 32S1H+ interfered with 32S and 33S
identification, leading to poor concentration recovery of 32S and 33S
15. Thus, element 34S in the 10% HCl (v/v) background
solution was selected for further study.
From the experimental data,
the primary mass 28Si appeared to have better concentration recovery
than the two isotopes (mass 29 and mass 30). The possible reason was that the
optimization of carrier gas to 0.65L/min (typically lower than 0.7L/min)
reduced solvent loading and decreased carbon accumulation avoiding 12C
interference, leading to less interference from
12C16O+
15,16.
After selected suitable
detection mass of Ca, S and Si, 17 elements’ identification focused on the
primary masses with minimum isotopes interference, except for element Pb 17-19.
The correction equation calculated the sum of 206Pb, 207Pb
and 208Pb responses in Table 3B was applied for 208Pb
identification to increase the sensitivity of detection. ICP-MS system
parameters for 21 elements were optimized based on the masses selected for 21
elements and stated in Table 3A and Table 3B.
Standard solution/sample
preparation optimization:
The matrix influence may lead
to background interference for elemental identification by ICP-MS. The suitable
matrix preparation can minimize the background impact. After several trials
comparing the identification results of trace metals by using deionized water
(DDW), 2%HCl v/v, and 10%HCl v/v as matrix, the solution preparation based on
the 10%HCl v/v matrix was regarded as the most suitable application for 21
elements identification. The rinse solution with 10%HCl v/v matrix between
samples could reduce the carry-over of samples on ICP-MS to perform the expected
trace element identification and quantitation.
Method validation via analysis
of spiked standard solutions in KCl buffer solution:
Spike stock solutions and test
solutions preparation:
All spiked stock standard
solutions used are stated in Table 1. For method development and validation,
one unspiked buffer solution and four different spiked concentrations were
prepared per Table 4A. After spiked stock solutions preparation, dilution factor
5 was performed to all unspiked and spiked stock solutions, before they were
introduced into ICP-MS. The stock solutions preparation and stock solutions
concentration are included in Table 4A and Table 4B. The test solution
preparation and sample amounts with dilution factor 5 in different day is shown
in Table 4C.
Table 4A. Unspiked and Spiked
Stock Solutions Preparation (unit: mL)
|
Solution ID |
Buffer (KCl pH 2 solution) |
EI-SPK-1A |
EI-SPK-2A |
EI- SPK-3 |
Std Hg 10ppm |
Std Ca 1000ppm |
Std S 1000ppm |
Std Si 1000ppm |
Final Volume with buffer |
|
Level 0 (Blank) |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
50 |
|
Quantitation Limit (QL) |
~25 |
0.025 |
0.030 |
0.030 |
0.030 |
0.020 |
0.030 |
0.030 |
50 |
|
Level-1 |
~25 |
0.050 |
0.050 |
0.050 |
0.050 |
0.030 |
0.050 |
0.050 |
50 |
|
Level 2 |
~25 |
0.200 |
0.200 |
0.200 |
0.100 |
0.050 |
0.075 |
0.075 |
50 |
|
Level-3 |
~25 |
0.450 |
0.650 |
0.450 |
0.150 |
0.075 |
0.100 |
0.100 |
50 |
Table 4B.
Concertation of Unspiked and Spiked Stock Solutions (unit: ng/mL, Dilution
Factor 5)
|
Elements |
Level 0 |
QL |
Level 1 |
Level 2 |
Level 3 |
|
Aluminum (Al) |
0 |
30 |
50 |
200 |
450 |
|
Silicon (Si) |
0 |
600 |
1000 |
1500 |
2000 |
|
Sulfur (S) |
0 |
600 |
1000 |
1500 |
2000 |
|
Calcium (Ca) |
0 |
400 |
600 |
1000 |
1500 |
|
Vanadium (V) |
0 |
25 |
50 |
200 |
450 |
|
Chromium (Cr) |
0 |
25 |
50 |
200 |
450 |
|
Manganese (Mn) |
0 |
25 |
50 |
200 |
450 |
|
Iron (Fe) |
0 |
30 |
50 |
200 |
650 |
|
Nickel (Ni) |
0 |
30 |
50 |
200 |
650 |
|
Cobalt (Co) |
0 |
12.5 |
25 |
100 |
225 |
|
Copper (Cu) |
0 |
25 |
50 |
200 |
450 |
|
Zinc (Zn) |
0 |
60 |
100 |
400 |
900 |
|
Arsenic (As) |
0 |
12.5 |
25 |
100 |
225 |
|
Selenium (Se) |
0 |
30 |
50 |
200 |
650 |
|
Molybdenum (Mo) |
0 |
25 |
50 |
200 |
450 |
|
Cadmium (Cd) |
0 |
12.5 |
25 |
100 |
225 |
|
Tin (Sn) |
0 |
60 |
100 |
400 |
900 |
|
Antimony (Sb) |
0 |
12.5 |
25 |
100 |
225 |
|
Barium (Ba) |
0 |
30 |
50 |
200 |
650 |
|
Mercury (Hg) |
0 |
6 |
10 |
20 |
30 |
|
Lead (Pb) |
0 |
12.5 |
25 |
100 |
225 |
Table 4C. Test
solution preparation
|
Test Solution ID |
Volume of SPK solution, mL |
Final Volume with Rinse (10% HCl), mL |
Dilution Factor |
Preparation Number |
|
|
1Day 1 |
2Day 2 |
||||
|
Level 0 |
5 |
25 |
5 |
3 |
3 |
|
QL |
5 |
25 |
5 |
3 |
0 |
|
Level 1 |
5 |
25 |
5 |
3 |
0 |
|
Level 2 |
5 |
25 |
5 |
6 |
6 |
|
Level 3 |
5 |
25 |
5 |
3 |
0 |
1 The data gained in day 1 was used for accuracy and precision, standard
and sample linearity, Matrix specificity, calibration range and QL.
2 Intermediate precision was
tested in two independent analyses.
Table 4D. Method
Performance Characteristics and Summary of Method Validation Data
|
Method Validation Parameters |
Experiment |
Expected Criteria per USP 233 and Ph. Eur. 2.4.20 20,21 |
Validation Data Summary
|
|
Accuracy and Precision of % Recovery |
Three replicate results for: · Level 0 · QL · Level 1 · Level 3 Six replicates for: · Level 2
Calculate the % recovery with respect to the theoretical concentration supplemented per Equation 1. |
Criteria of Mean recoveries: 70% -150%, RSD: ≤20%.
No criteria for 95% Confidence Interval |
Mean recoveries: 93% ≤ QL ≤ 112%, 92% ≤ Level 1 ≤ 110% 90% ≤ Level 2 ≤ 110% 90% ≤ Level 3 ≤ 108%
%RSD: 0.870% ≤ QL ≤ 5.433% 0.228% ≤ Level 1 ≤3.591% 0.314% ≤ Level 2 ≤3.536% 0.166% ≤ Level 3 ≤2.205% |
|
Precision – Repeatability |
Six replicate analyses (n=6) of Level 2 in day 1 and day 2. Calculate individual and mean concentration, %RSD from the 6 replicate corrected measurements |
RSD ≤ 20% for each analysis
|
0.314% ≤ RSD for analysis1 ≤ 3.536%
0.669% ≤ RSD for analysis 2 ≤ 3.366% |
|
Precision – Intermediate (Ruggedness) |
Six replicate analyses of Level 2 from each analysis. Calculate mean concentration corrected from the 12 replicate (n=12) measurements %RSD and RPD |
RSD: ≤25% RPD ≤ 25% |
0.056% ≤ RPD ≤ 4.286% 0.678% ≤ RSD ≤ 3.403% |
|
Matrix Specificity |
Accuracy and precision requirements |
Meet the Accuracy and Precision Requirements |
Met the Accuracy and Precision Requirements |
|
Range of the Method |
The lowest supplemented to the highest supplemented concentrations determined |
Meet the accuracy, precision and linearity requirements. |
Range of 21 elements (QL – Level 3) can refer Table 5B |
|
Quantitation limit (QL) |
The lowest amount of analyte in a sample that can be determined with acceptable precision and accuracy under the stated experimental conditions |
Meet the Accuracy and Precision Requirements. |
QL of each element can refer Table 5B |
|
Standard and Sample Linearity |
Linear regression from the analyses performed for the supplemented Test Solutions corrected for the Level 0 concentrations determined in the accuracy and precision experiment |
Correlation coefficient r ≥ 0.995 Report Slope and Intercept |
Sample linearity: Correlation coefficient 0.997≤ r ≤ 1.000 Slope and Intercept can refer Table 5B
Standard linearity: Correlation coefficient 0.9999≤ r ≤ 1.0000 Slope and Intercept can refer Fig 1 |
Method validation reference
and criteria expected:
The method validation
procedure and parameters referred Category II Quantitative method for Impurity
and the “Alternative procedure validation” allowance described in
USP<233> and Ph. Eur. 2.4.2020,21. The validation parameters,
experiment designs and the expected criteria following the two regulations are
detailed in Table 4D.
Experiment design and data
analysis of accuracy and precision:
The working standard (WS)
solutions were prepared per Table 2A and tested for 21 elements standard
linearity calibration. While triplicate of unspiked solution (Level 0),
Quantitation Limit (QL), Level 1 and Level 3 spiked solution and six replicates
of Level 2 spiked solution with dilution factor 5 as the description in Table
4C for 21 elements were tested. The sample concentration data gained for each
element was used for accuracy and precision calculation.
The experimental concentration
was determined by comparison of the experimental test solution response to the
standard response. Accuracy as analytical recovery was calculated per Equation
1. If the average of Level 0 (unspiked) concentration was less than zero, then
zero was used for Level 0 concentrations for the percent recovery calculation.
Mean % Recovery of the percent recoveries was calculated for the applicable
test solutions. The % RSD of the percent recoveries and experimental
concentration of the spiked concentration that were not corrected to the
inherent (Level 0) were calculated as repeatability.
Accuracy as % Recovery
Recovery (%) =
Experiemental Conc. Spiked (ng
⁄mL) - Mean of Experimental Conc. Unspiked (ng⁄mL)/Theoretical
Conc. Spiked (ng⁄mL) × 100
%
Equation 1
Experiment design and data
analysis of method robustness:
Two independent analyses for
21 elements determination were conducted to estimate the method robustness. Six
replicate analyses of Level 2 from each analysis (n=6) was analyzed.
The repeatability and
intermediate precision (robustness) were calculated from the experimental
results via two analytical runs (six replicate analyses performed on two
independent analyses). Repeatability was calculated from experimental
concentration of the spiked concentration not corrected to the inherent (see
accuracy section above).
Robustness were calculated
from the experimental results of the test solutions from two analytical runs
(n=12). The spiked solution concentrations were corrected to Level 0, if the
mean of Level 0 concentrations were higher than zero. The repeatability percent
relative standard deviation (% RSD) of the six determinations for the corrected
spiked solutions from Level 0 on each analysis was calculated. The intermediate
precision relative percent difference (RPD) from each determination (12 data
from two analyses) for the corrected spiked solutions from each run was
calculated.
The RPD was calculated using
Equation 2 and the absolute difference between the two results. Mean1
was the mean result from precision experiment and Mean2 was the
result from the intermediate precision analysis only.
RPD = (Corrected Sample
concentration Mean2 - Corrected Sample concentration Mean1)/(Corrected
Sample concentration Mean1) ×
100
Equation 2
Experiment design and data
analysis of standard and sample linearity:
The experimental analysis of
working standard solutions and the spiked solutions prepared for accuracy and
precision tests were used for standard and sample linearity calculation,
respectively. Standard linearity was calculated per the linear regression of
the experimental versus the theoretical concentration of five Working Standard
(WS0, WS1, WS2, WS3 and WS4) solutions for each element via calculating by
ICP-MS Masshunter software.
Sample linearity was assessed
from the linear regression of the experimental versus the theoretical
concentration of the test solutions (QL, Level 1, Level 2 and Level 3) for each
element. The mean of Level 0 test solution concentrations was subtracted from
the spiked test solution concentrations, if it was greater than zero. The
correlation coefficient (r), slope and y-intercept were calculated and
reported.
Experiment design and data
analysis of matrix specificity, QL and calibration range based on accuracy and
precision:
The matrix specificity of the
procedure for determining target elements in the presence of the matrix was
demonstrated by the accuracy and precision determined in the test solutions
spiked with target elements. The method calibration range was defined as the
lowest and highest spiked concentrations for the target elements that produce
acceptable accuracy, precision, and linearity results. The QL was the lowest
spiked concentration (ng/mL) of the target elements that produced acceptable
accuracy and precision results.
RESULTS AND DISCUSSION:
The main aspects of method
development included elemental mass selection, ionization and detection
optimization of ICP-MS without auxiliaries, exploration of the test solution
preparation for trace elements Ca, S and Si as well as 18 trace elements22.
The purpose was to maximize the sensitivity and minimize the atomic, ions and
background interferences. The following sections validated the method for 21
trace elements analysis with QL in nanogram per milliliter from the extractable
KCl buffer solution.
Standard linearity
calculation:
Five calibration points
included one blank and 4 different concentration standards were studied for
standard linear regression calculation (response ratio vs. theoretical
concentration) by Agilent ICP-MS Masshunter software. The linear regression
calibration results in Fig 1 showed that the linearity calibration (r value) of
21 elements was in the range of 0.9999-1.0000. The good linearity of standard
solutions illustrates the new developed method is capable to analyze 21
elements in nanogram per milliliter level.
A: Standard linearity calibration curve of 27Al
B:
Standard linearity calibration curve of 28Si
C: Standard linearity calibration curve of 34S
D:
Standard linearity calibration curve of 44Ca
E: Standard linearity calibration curve of 51V
F:
Standard linearity calibration curve of 52Cr
G: Standard linearity calibration curve of 55Mn
H:
Standard linearity calibration curve of 56Fe
I: Standard linearity calibration curve of 58Ni
J:
Standard linearity calibration curve of 59Co
K: Standard linearity calibration curve of 63Cu
L:
Standard linearity calibration curve of 66Zn
M: Standard linearity calibration curve of 75As
N:
Standard linearity calibration curve of 78Se
O: Standard linearity calibration curve of 95Mo
P:
Standard linearity calibration curve of 111Cd
Q: Standard linearity calibration curve of 118Sn
R:
Standard linearity calibration curve of 121Sb
S: Standard linearity calibration curve of 137Ba
T:
Standard linearity calibration curve of 201Hg
U: Standard linearity calibration curve of 208Pb.
Fig 1. Standard Linearity
Calibration Curve of Each Element by Agilent ICP-MS Masshunter software
*Ratio: the signal intensity of each element; Conc. (ng/ml): theoretical
concentration of each element in standard solution
DL (ng/ml): Detection Limit, BEC (ng/ml): Background Equivalent
Concentration
The big challenge of this
method development was considered as the identification and quantification of 44Ca,
34S and 28Si, because of the ion and matrix interferences15.
Based on the linearity regression curve in Fig 1, the standard linearity
regression r values of the three elements were higher than 0.9999. Regarding 28Si
determination, 14N2+ removal mostly
contributed less ions interference on 28Si identification under the
analytical method. The optimization of carrier gas reduced solvent loading and
decreased carbon accumulation avoiding 12C16O+
interference15,16. After carrier gas optimization, the less
abundance of isotopes 29Si and 30Si have low response.
Hence, 28Si had the most significant response signal under the
current analytical parameters.
In addition, the fitting curve
of element 34S (y = 0.0013* x + 0.6381, r=0.9999) had a very good
linearity regression calibration. But considering the linearity calibration
data shown in Fig 1C, ICP-MS instrumental Detection Limit (DL) showed
32.96ng/mL. The high DL of 34S might came from interference of
isotopic ions e.g. 16O18O+ and 17O2+
in plasma, 16O17O1H+ and 33S1H+
in acidic matrix9,15. The isotopic and background interferences were
of the inherent characteristics of ICP-MS and analytes, e.g. O2 in
plasma and H+ in acidic matrix. With the newly developed method,
this methodological DL of element 34S and 28Si reached
100 ng/mL to meet the needs of extractable elements’ toxic assessment.
Regarding 44Ca
linearity regression calibration (y=0.0280 * x +0.0883), the quantification did
not show high background noise and ions interference considering the low
intercept in Y-axis in Fig 1D. The methodological DL of 44Ca was
15ng/mL by using this new developed method. Except for the standard linearity
of three most difficult elements, the standard linearity of other 18 elements
were calibrated and performed good linearity regression. In Fig 1, the standard
linearity regression r ranged from 0.9999-1.0000 with very low intercepts for
the other 18 elements. That indicated that all 21 elements could be identified
and quantified by the optimized ICP-MS method with the minimized interference
from ions and atomic background on 21 elemental analysis under current ICP-MS
status.
Analysis of 21 elements in pH
2 KCl buffer solutions spiked with known concentration standard solutions:
Selection the buffer solution
and dilution factor performed:
Aiming the elemental
extraction from packaging materials, on one hand, different extraction
solutions need to maximize the extraction efficiency for individual element. On
the other hand, the selected extraction solvent needs to carry the least matrix
interference on elemental analysis. Based on these two principles, the most
common extraction solvents for various packaging materials are of pH 2-12
water/salt buffer1.
In this study, KCl pH 2 buffer
solution was considered to have minimum matrix influence with cation in
salt/water buffer for trace elemental analysis by ICP-MS and mimic the worst
case of drug products effected the extractable elemental impurities from
packaging systems. Certainly, alkaline buffer and organic extract solvents are
also applied in extractable studies 1. However, due to the acidification
of analytes/samples, alkaline buffer solvent is not suitable for current
analysis by ICP-MS while organic solvent leads to unstable plasma of ICP-MS.
KCl pH 2 buffer solution
spiked with the known concentration standard solutions was selected to verify
the developed method’s accuracy and precision. During the initial analysis of
spiked KCl pH 2 buffer solution, the response recovery of ISTD elements 89Y,
115In and 238U against the corresponding response in the
blank working standard solution (WS0) were separately 288%, 255% and 161%, when
the spike test solutions were directly injected into ICP-MS without any
dilution. This illustrated that the impact of high concentration salt ions in
KCl pH 2 buffer solution on ISTD analysis was strong. The response of ISTD
relies on ISTD elements’ first ionization potential (IP). ISTD elements with
high first IPs can be suppressed by the presence of easily ionized matrix
element(s) at a relatively high concentration such as K+ in the
buffer solution 23,24. Therefore, the effective way to reduce the
relative suppression is appropriate dilution of the high concentration salts.
The ISTD element recovery was
compared by applying different dilution factor on test solutions. The direct
sample introduction into ICP-MS resulted in over 250% recovery of 89Y
and 115In. On the opposite, the low recovery of 89Y, 115In
and 238U was between 58%-69% by applying a dilution factor 10 on
spiked KCl pH 2 buffer. Hence, dilution factor 5 was suitable for trace element
analysis in the spike buffer solution.
Accuracy and precision of the
method:
The accuracy and precision of
elements in KCl pH 2 buffer solvent with spiked known concentration standard
solutions analyzed by the developed method are reported in Table 5A. Within the
calibration range of each element, the identification and quantification of 21
trace elements met excellent accuracy and precision. In all spiked
concentration levels, the concentration recovery of 21 elements ranged from
90%-112% while the relative standard deviation (RSD) of recovery was below
5.433%. The elemental impurities analysis for 21 trace elements could well
follow the compendia criteria of USP 233 and Ph. Eur. 2.4.20, which required
accuracy of spike recovery 70%–150% for the mean of three replicate
preparations at each concentration and precision Not More Than (NMT) 20% of
each target element in concentration levels20,21. The quantitation
limit and calibration range of 21 trace elements by this method are shown in
Table 5B.
Table 5A. 1Accuracy
and 2Precision of Each Element for Unspiked and Spiked Solutions
Analysis (Dilution Factor 5)
|
Ele ment |
Level 0 |
|
QL |
|
||
|
Mean Conc. ng/mL |
Thermotical Conc., ng/mL |
Avg. Conc., ng/mL |
Accuracy, % |
Precision, % |
Thermotical Conc, ng/mL |
|
|
27Al |
1.34 |
30 |
35.08 |
112 |
2.54 |
50 |
|
28Si |
0 |
600 |
619.58 |
103 |
1.96 |
1000 |
|
34S |
111.65 |
600 |
751.77 |
107 |
5.43 |
1000 |
|
44Ca |
3.63 |
400 |
449.84 |
112 |
0.87 |
600 |
|
51V |
0 |
25 |
25.66 |
103 |
2.60 |
50 |
|
52Cr |
0 |
25 |
25.04 |
100 |
2.08 |
50 |
|
55Mn |
0.03 |
25 |
25.33 |
101 |
2.67 |
50 |
|
56Fe |
0.53 |
30 |
30.54 |
100 |
2.11 |
50 |
|
58Ni |
0 |
30 |
29.35 |
98 |
1.68 |
50 |
|
59Co |
0.01 |
12.5 |
12.69 |
101 |
2.01 |
25 |
|
63Cu |
0.02 |
25 |
24.23 |
97 |
2.04 |
50 |
|
66Zn |
0.01 |
60 |
55.65 |
93 |
1.50 |
100 |
|
75As |
0 |
12.5 |
13.03 |
104 |
1.45 |
25 |
|
78Se |
0.05 |
30 |
29.27 |
97 |
1.06 |
50 |
|
95Mo |
0 |
25 |
24.73 |
99 |
1.55 |
50 |
|
111Cd |
0.14 |
12.5 |
12.20 |
96 |
2.19 |
25 |
|
118Sn |
0.03 |
60 |
61.50 |
102 |
1.05 |
100 |
|
121Sb |
0.16 |
12.5 |
12.91 |
102 |
1.35 |
25 |
|
137Ba |
0.13 |
30 |
31.28 |
104 |
2.34 |
50 |
|
201Hg |
0.53 |
6 |
6.36 |
97 |
1.54 |
10 |
|
208Pb |
1.51 |
12.5 |
14.03 |
100 |
1.33 |
25 |
Table 5A. Continued
|
Elem ent |
Level 1 |
|
Level 2 |
|
Level 3 |
||||||
|
Avg. Conc., ng/mL |
Accuracy, % |
Precision, % |
Thermotical Conc., ng/mL |
Avg. Conc., ng/mL |
Accuracy, % |
Precision, % |
Thermotical Conc., ng/mL |
Avg. Conc., ng/mL |
Accuracy, % |
Precision, % |
|
|
27Al |
56.50 |
110 |
1.87 |
200 |
220.62 |
110 |
0.76 |
450 |
484.83 |
107 |
0.93 |
|
28Si |
1055.97 |
106 |
1.97 |
1500 |
1640.00 |
109 |
0.58 |
2000 |
2167.80 |
108 |
0.58 |
|
34S |
1148.74 |
104 |
3.59 |
1500 |
1735.10 |
108 |
3.54 |
2000 |
2243.27 |
107 |
0.43 |
|
44Ca |
665.45 |
110 |
0.23 |
1000 |
1090.72 |
109 |
1.43 |
1500 |
1617.43 |
108 |
2.21 |
|
51V |
51.33 |
103 |
1.72 |
200 |
207.76 |
104 |
0.58 |
450 |
470.22 |
104 |
0.58 |
|
52Cr |
49.42 |
99 |
1.65 |
200 |
199.33 |
100 |
0.50 |
450 |
457.20 |
102 |
1.13 |
|
55Mn |
49.40 |
99 |
1.33 |
200 |
197.67 |
99 |
0.44 |
450 |
445.27 |
99 |
0.60 |
|
56Fe |
49.49 |
98 |
1.14 |
200 |
196.09 |
98 |
0.40 |
650 |
656.33 |
101 |
1.53 |
|
58Ni |
47.92 |
96 |
1.07 |
200 |
191.95 |
96 |
0.47 |
650 |
626.06 |
96 |
1.12 |
|
59Co |
24.89 |
100 |
1.23 |
100 |
99.29 |
99 |
0.57 |
225 |
223.42 |
99 |
0.52 |
|
63Cu |
47.66 |
95 |
1.09 |
200 |
190.54 |
95 |
0.63 |
450 |
432.73 |
96 |
1.38 |
|
66Zn |
91.61 |
92 |
1.64 |
400 |
361.22 |
90 |
0.70 |
900 |
809.27 |
90 |
0.70 |
|
75As |
25.86 |
103 |
2.13 |
100 |
103.48 |
103 |
0.47 |
225 |
230.16 |
102 |
0.69 |
|
78Se |
48.30 |
96 |
0.90 |
200 |
191.04 |
95 |
0.59 |
650 |
611.93 |
94 |
0.17 |
|
95Mo |
48.74 |
97 |
0.72 |
200 |
193.70 |
97 |
0.53 |
450 |
435.98 |
97 |
0.54 |
|
111Cd |
23.93 |
95 |
0.77 |
100 |
94.18 |
94 |
0.39 |
225 |
207.69 |
92 |
0.33 |
|
118Sn |
100.81 |
101 |
0.49 |
400 |
396.65 |
99 |
0.62 |
900 |
890.70 |
99 |
1.15 |
|
121Sb |
25.49 |
101 |
0.90 |
100 |
100.38 |
100 |
0.70 |
225 |
221.66 |
98 |
0.56 |
|
137Ba |
51.62 |
103 |
1.05 |
200 |
201.97 |
101 |
0.74 |
650 |
645.22 |
99 |
0.38 |
|
201Hg |
10.47 |
99 |
1.63 |
20 |
21.00 |
102 |
1.02 |
30 |
30.80 |
101 |
1.22 |
|
208Pb |
26.08 |
98 |
1.35 |
100 |
99.40 |
98 |
0.47 |
225 |
220.62 |
97 |
0.24 |
1 Accuracy as concentration recovery was calculated per
Equation 1.
2 Precision was
calculated as %RSD of the percent recoveries based on accuracy.
Table 5B. Calibration of Sample
Linearity, QL and Calibration Range of Each Element for Unspiked and Spiked Solutions
Analysis (Dilution Factor 5)
|
Element |
Slope |
Intercept |
Correlation Coefficient (r) |
Calibration Range (ng/ml) |
QL (ng/ml) |
|
27Al |
1.07 |
2.85 |
0.9999 |
30-450 |
30 |
|
28Si |
1.12 |
-47.68 |
0.9995 |
600-2000 |
600 |
|
34S |
1.08 |
-16.00 |
0.9967 |
600-2000 |
600 |
|
44Ca |
1.06 |
24.14 |
0.9992 |
400-1500 |
400 |
|
51V |
1.05 |
-1.02 |
1.0000 |
25-450 |
25 |
|
52Cr |
1.02 |
-2.10 |
0.9999 |
25-450 |
25 |
|
55Mn |
0.99 |
0.13 |
1.0000 |
25-450 |
25 |
|
56Fe |
1.01 |
-3.41 |
0.9998 |
30-650 |
30 |
|
58Ni |
0.96 |
-0.22 |
0.9999 |
30-650 |
30 |
|
59Co |
0.99 |
0.14 |
1.0000 |
12.5-225 |
12.5 |
|
63Cu |
0.96 |
-0.73 |
0.9999 |
25-450 |
25 |
|
66Zn |
0.90 |
2.03 |
1.0000 |
60-900 |
60 |
|
75As |
1.02 |
0.65 |
0.9999 |
12.5-225 |
12.5 |
|
78Se |
0.94 |
2.00 |
1.0000 |
30-650 |
30 |
|
95Mo |
0.97 |
0.34 |
1.0000 |
25-450 |
25 |
|
111Cd |
0.92 |
1.18 |
0.9999 |
12.5-225 |
12.5 |
|
118Sn |
0.99 |
2.00 |
0.9999 |
60-900 |
60 |
|
121Sb |
0.98 |
1.13 |
0.9999 |
12.5-225 |
12.5 |
|
137Ba |
0.99 |
2.69 |
1.0000 |
30-650 |
30 |
|
201Hg |
1.02 |
-0.21 |
0.9995 |
6.0-30 |
6.0 |
|
208Pb |
0.97 |
0.43 |
1.0000 |
12.5-225 |
12.5 |
Facing the analytical
challenge for Ca, S and Si by ICP-MS, the mean concentration recovery of 44Ca
elements were from 108% -112% in different spike levels with precision less
than 2.21%. The concentration recovery of 28Si was 103%-109% and
recovery %RSD of it was lower than 2% in different spike concentrations,
respectively. The reduced matrix interference (e.g. removed HNO3 in
the background and avoiding 12C16O+
accumulation by optimization of carrier gas, etc.) overcame the difficulty of
the two elements’ analysis.
During the method development,
the recovery %RSD of 34S was initially around 20% (the upper limit
of %RSD criteria) for all spiked levels within the calibration range, even if
the concentration recovery was between 90-110%. The reason of high %RSD for
element 34S was the high background noise caused by the inherent
character and the acidic matrix per the standard linearity calibration data
showing in Fig 1C. The high background noise led to high instrumental DL
32.96ng/mL for element 34S in standard solutions. In buffer solution
with high concentration salts, the high total dissolved solids (TDS) caused a
downward drift of analyte signal25. Accordingly, high %RSD appeared
in 34S analysis for all spiked levels.
In the end, 21 elements in one
single sample analyzed by two sublists in one analytical sequence worked well
to decrease the recovery %RSD of 34S. One sublist included 20
elements and excluded element S while another sublist analyzed element S only
in the analyte list. Since the oxygen interference existing in plasma, isotopic
interference and signal downward drift caused by the high TDS in high
concentration salts contributed the low ionization of sulfur in argon plasma,
the sublist only focused on S analysis might increase the ionization of sulfur
to some degree by increasing plasma robust without other elements existed in
the same analyte
list 15,
25-27. As a result, one sample with 21 trace elements was analyzed twice
by two sublits in one sequence to obtain ideal accuracy and precision results for
all targeted elements.
Matrix Specificity,
Calibration range, QL and Sample linearity of the method:
Matrix specificity was
demonstrated by the accuracy and precision of the method. As the stated in the
section above, the accuracy and precision of the method had very good
performance in the defined calibration range. The matrix interference was
acceptable without significantly effect on this analytical method.
In Table 5B, calibration range
and QL of each element in buffer solution were listed. After dilution by factor
5, the concentration of each element in buffer solution was within the standard
calibration range. The lowest concentration within the sample calibration range
is QL of the targeted element. As for 28Si and 34S, QL of
them in buffer solution was 600ng/mL and introduced to 120ng/mL to ICP-MS with
dilution factor 5. Due to the capability limitation of current ‘ordinary’
ICP-MS without any auxiliary, it is not easy to reach a very low QL. The
typical example is element 34S with high instrumental DL 32.96ng/mL
caused by the background noise of instrument. Current QL of all elements
studied in KCl buffer solution are sufficient for toxic assessment for
packaging material in the future.
The sample linearity
calibration of each element, which Correlation Coefficient (r) was in the range
of 0.9967 – 1.0000 under the method calibration range after dilution by factor
5, met the expected criteria (r ≥ 0.995) of USP 233 and Ph. Eur. 2.4.20 20,21.
The data could be referred in Table 5B.
Robustness (Repeatability and
Intermediate Precision) of the method:
The method robustness
reflected via the repeatability and intermediate precision. The robustness of
method was validated and the data was reported in Table 5C. From this method
validation data in Table 5C, the repeatability of each analysis (%RSD, n=6) for
all 21 elements was lower than 3.54%. Meanwhile, the repeatability of two
analysis (%RSD, n=12) for 21 elements was lower than 3.40%. Moreover, RPD of
two analysis for 21 elements was lower than 1.64%. The good performance of the
method completely followed the criteria USP 233 and Ph. Eur. 2.4.20 dominates
that the new developed method is capable to provide reliable data for 21
elements’ identification and quantification, even in a slightly varied
condition 20,21.
Table 5C. Repeatability and
Intermediate Precision of Each Element for Unspiked and Spiked
Solutions Analysis (Dilution Factor 5)
|
Ele ment |
Analysis 1, Corrected Level 2 (n=6) |
Analysis 2, Corrected Level 2 (n=6) |
Analysis 1 & Analysis 2 Corrected Level 2 (n=12) |
||||
|
|
1Mean Conc.,(ng/mL) |
%RSD |
1Mean Conc.,(ng/mL) |
%RSD |
1Mean Conc.,(ng/mL) |
%RSD |
RPD |
|
27Al |
219.27 |
0.76 |
215.78 |
1.76 |
217.53 |
1.53 |
1.59 |
|
28Si |
1640.00 |
0.58 |
1626.81 |
1.31 |
1633.40 |
1.05 |
0.80 |
|
34S |
1623.46 |
3.54 |
1596.88 |
3.37 |
1610.17 |
3.40 |
1.64 |
|
44Ca |
1087.09 |
1.43 |
1097.19 |
0.76 |
1092.14 |
1.19 |
0.93 |
|
51V |
207.76 |
0.58 |
207.26 |
1.34 |
207.51 |
1.00 |
0.24 |
|
52Cr |
199.33 |
0.50 |
199.04 |
1.41 |
199.18 |
1.01 |
0.15 |
|
55Mn |
197.65 |
0.44 |
200.01 |
1.49 |
198.83 |
1.22 |
1.19 |
|
56Fe |
195.56 |
0.40 |
194.06 |
1.51 |
194.81 |
1.12 |
0.77 |
|
58Ni |
191.95 |
0.47 |
189.76 |
1.12 |
190.85 |
1.01 |
1.14 |
|
59Co |
99.28 |
0.57 |
98.49 |
1.12 |
98.89 |
0.94 |
0.80 |
|
63Cu |
190.52 |
0.63 |
190.73 |
1.48 |
190.63 |
1.09 |
0.11 |
|
66Zn |
361.21 |
0.70 |
360.58 |
1.72 |
360.89 |
1.26 |
0.17 |
|
75As |
103.48 |
0.47 |
103.74 |
1.44 |
103.61 |
1.03 |
0.25 |
|
78Se |
190.98 |
0.59 |
189.56 |
1.67 |
190.27 |
1.25 |
0.74 |
|
95Mo |
193.70 |
0.53 |
194.99 |
1.02 |
194.34 |
0.85 |
0.66 |
|
111Cd |
94.04 |
0.39 |
93.46 |
0.87 |
93.75 |
0.72 |
0.62 |
|
118Sn |
396.62 |
0.62 |
397.87 |
1.19 |
397.24 |
0.92 |
0.31 |
|
121Sb |
100.22 |
0.70 |
100.17 |
1.24 |
100.19 |
0.96 |
0.06 |
|
137Ba |
201.85 |
0.74 |
199.50 |
0.90 |
200.67 |
0.99 |
1.16 |
|
201Hg |
20.48 |
1.02 |
20.23 |
1.71 |
20.35 |
1.48 |
1.22 |
|
208Pb |
97.89 |
0.47 |
97.15 |
0.67 |
97.52 |
0.68 |
0.76 |
1Corrected supplemented solution
concentrations for the unsupplemented concentrations (Level 0), when the
average unsupplemented concentrations is higher than 0 (zero)
CONCLUSIONS:
The new developed and
validated method can simultaneously identify and quantify 21 trace elements
from pharmaceutical packaging materials with excellent performance by using
ICP-MS without any auxiliary in a single method. The application of this method
in acidic extractable solvent with the spiked known concentration standard
solutions demonstrated successful accuracy and precision of all 21 elements in
nanogram per milliliter level. By meeting US and European compendia criteria,
the identification and quantification of elemental impurities from packaging
materials provide the essential information of risk assessment for the
packaging materials and relevant final drug products. The success of this
method convinces us that the trace elements analysis from the packaging
materials by a simple method without auxiliary and with relatively low
operational cost in pharmaceutical industry is feasible and reliable. This
study focused on the method development and validation for 21elements
determination. But it was limited to method applications in a wider range of
extractable solvents and real samples. Further research should focus on method
applications in various package materials to gain more valuable data for risk
assessment of pharmaceutical packaging materials.
ACKNOWLEDGEMENT:
It is grateful that this work
has been greatly supported by Shared Chemistry Team in R&D Center, Baxter
Suzhou, China.
DECLARATION OF CONFLICTING INTERESTS:
The authors declared no
potential conflicts of interest with respect to the research, authorship,
and/or publication of this article.
REFERENCES:
1.
Jenke, D. R.; Stults, C. L. M.; Paskiet, D. M.; et al. Materials
in Manufacturing and Packaging Systems as Sources of Elemental Impurities in
Packaged Drug Products: A Literature Review, PDA J Pharm. Sci. and Tech.
2015, 69, 1-48. DOI:10.5731/pdajpst.2015.01005
2.
The International Council for Harmonisation of Technical
Requirements for Pharmaceuticals for Human Use (ICH). Guideline for Elemental
Impurities. Q3D, version dated December 2014. R1 Adopted on 22 March 2019
https://www.ich.org/products/guidelines/quality/article/quality-guidelines.html
3.
Saini S.; Nanda S.; Dhiman A. Elemental Analysis in Piper betle Linn.
and Jatropha gossypifolia Linn. leaves: Biosafety Studies. Research J.
Pharm. and Tech.,2018,11(11), 5078-5082. DOI:
10.5958/0974-360X.2018.00926.5
4.
Lisboa, M. T.; Clasen, C. D.; de Sousa Vellar, D. C.; et al. An
Easy and Fast Procedure for the Determination of Ca, Cu, Fe, Mn, Mg, Na, K and
Si in Biodiesel by ICP OES Using Emulsification as Sample Preparation Strategy. J. Braz. Chem.
Soc., 2014, 25(1), 143-151. http://dx.doi.org/10.5935/0103-5053.20130280
5.
Sarojam, P., PerkinElmer, Inc. Analysis of Wastewater for Metals using
ICP-OES. ICP-Optical Emission Spectroscopy application note.
https://docplayer.net/38896524-Analysis-of-wastewater-for-metals-using-icp-oes-icp-optical-emission-spectroscopy-application-note-authors-praveen-sarojam-ph-d.html
6.
Ramesh Kumar K.A.; Pavithra G.M.; Syed J.; et. al. Elemental
analysis and anticariogenic activity of flowers of Calycopteris floribunda,
Humboldtia brunonis and Kydia calycina. Asian Journal of Research in
Chemistry. 2013, 6(7). 623-627.
7.
Shukla, A.; Choudhary, A.; Shukla Rishi, K. Studies on Physico-chemical
Characteristics, Heavy Metals and Minerals components of leaves of Boehmeria
rugulosa Wedd. Research J. Pharm. and Tech. 2019, 5 (12). 2093-2097.
DOI: 10.5958/0974-360X.2019.00347.0
8.
Patnaik, P. S.; Ramanaiah M.; Ramaraju B. Quantitative Determination of
Essential and Trace Element Content of Some Medicinal Plants by ICP-MS
Technique. Research J. Pharm. and Tech. 2019, 4 (12). 1595-1600. DOI:
10.5958/0974-360X.2019.00265.8
9.
Mason, P. R. D.; Kaspers, K., van Bergen. M. J. Determination of sulfur
isotope ratios and concentrations in water samples using ICP-MS incorporating
hexapole ion optics, J. Anal. At. Spectrom., 1999, 14, 1067–1074. DOI: 10.1039/A902037C
10. Klemens, P.;
Heumann, K. G. Development of an ICP–HRIDMS method for accurate determination
of traces of silicon in biological and clinical samples, Fresenius J. Anal.
Chem., 2001, 371, 758–763. DOI 10.1007/s002160100996
11. Wollenweber, D.,
Straßburg, S., Wünsch, G. Determination of Li, Na, Mg, K, Ca and Fe with ICP-MS
using cold plasma conditions, Fresenius J. Anal. Chem., 1999, 364
(5),433–437. https://rd.springer.com/content/pdf/10.1007%2Fs002160051363.pdf
12. Evans, E. H.;
Wolff, J. C.; Eckers, C. Sulfur-Specific Detection of
Impurities in Cimetidine Drug Substance Using Liquid Chromatography Coupled to
High Resolution Inductively Coupled Plasma Mass Spectrometry and Electrospray
Mass Spectrometry. Anal. Chem. 2001, 73, 4722-4728. 10.1021/ac0103017
13. Chen, K. L.;
Jiang, S. J. Determination of calcium, iron and zinc in milk powder by reaction
cell inductively coupled plasma mass spectrometry. Analytica Chimica Acta.
2002, 470(2), 223–228. https://doi.org/10.1016/S0003-2670(02)00768-7
14. Thermo Fisher
Scientific, (Bremen) GmbH, Semi-Quantitative Environmental Analysis using the
XSERIES 2 ICP-MS, application note 40740.
https://static.thermoscientific.com/images/D00649~.pdf
15. May, T. W.;
Wiedmeyer, R. H. A Table of Polyatomic Interferences in ICP-MS, Atomic
Spectrometry, 1998, 19(5), 150-155. https://www.perkinelmer.com/CMSResources/Images/44-74379ATL_TableOfPolyatomicInterferences.pdf
16. Talbott, J.
Improving sensitivity, flexibility, and capabilities with Agilent’s ICP-MS and
Triple Quadrupole ICPMS with Tandem MS/MS capabilities For the Environmental,
Petrochem and Semiconductor Industries. Presentation in Houston Symposium
and Tradeshow, June 27, 2017.
https://www.agilent.com/cs/library/slidepresentation/public/ASTS-2013_7700_ICP-MS_and_8800_ICP_TripleQuad.pdf
17. Azize S. W. 5.
Study of Heavy Metals and their effects on Oxidant / Antioxidant Status in
Workers of fuel Station in Hilla city-Iraq. Research J. Pharm. and Tech.
2018, 11(1), 312-316. DOI: 10.5958/0974-360X.2018.00057.4
18. Kadhim Z. J.;
Hamaza M. F.; Khethi M. T. Study of Distribution of Trace Elements in Tigris
River in Baghdad City. Asian J. Research Chem. 2013,6(9), 811-813.
19. Al-Fartusie F. S.;
Mohssan S. N.; Risan F. A., et. al. Evaluation of Trace Elements and
Heavy Metals in Schizophrenic Patients in Iraq. Research J. Pharm. and Tech.
2019, 12(1), 185-191. DOI: 10.5958/0974-360X.2019.00034.9
20. USP <233>
Elemental Impurities—Procedures. Effective Date :01 May 2018.
https://www.uspnf.com/
21. Ph. Eur. (European
Pharmacopoeia) 9th Edition 2018 (9.6) <2.4.20> “Determination
of metal catalyst or metal reagent residues”. Official from July, 2018. http://online.pheur.org/EN/entry.htm#
22. Mujoriya, R. Z. Analytical
Method Development and Validation of Pharmaceutical Technology: An Overview. Research
J. Pharma. Dosage Forms and Tech. 2013; 5(4): 213-220
23. Gajek R.; Barley
F.; She J. W. Determination of essential and toxic metals in blood by ICP-MS
using calibration in synthetic matrix. Anal. Methods, 2013, 5,
2193-2202. DOI: 10.1039/c3ay26036d
24. Mary Saroja P.;
Immanuel G.; Allen Gnana Raj G. Trace Elements in Marine Macro algae
(Gracilaria edulis) on the Southern Coast of Tamil Nadu, India. Asian J.
Research Chem. 2014,7(3), 295-299.
25. Agatemor, C.,
Beauchemin D. Matrix effects in inductively coupled plasma mass spectrometry: A
review. Analytica Chimica Acta. 2011, 706
66–83. https://doi.org/10.1016/j.aca.2011.08.027
26. Fernández-Turiel,
J. L.; Llorens, J. F.; López-Vera F.; etc. Strategy for water analysis using
ICP-MS. Fresenius' Journal of Analytical
Chemistry. 2000, 368 (6), 601–606. https://doi.org/10.1007/s002160000552
27. Rajeshwari B. M.
Mass Attenuation Co-Efficient Inorganic Compound. Asian J. Research Chem. 2016,
9(12) 664-668. DOI: 10.5958/0974-4150.2016.00091.2
Received on 18.11.2019 Modified on 21.12.2019
Accepted on 11.01.2020
©Asian Pharma Press All Right Reserved
Asian J. Pharm. Ana. 2020; 10(2):51-66.
DOI: 10.5958/2231-5675.2020.00011.3