Orthogonal pre-use and post-use efficiency testing for single-use anion exchange chromatography

Journal of Chromatography A 1654 (2021) 462445

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Orthogonal pre-use and post-use efficiency testing for single-use
anion exchange chromatography
Jonathan F. Hester a,∗, Xinran Lu a, Jacob D. Calhoun a, Rebecca A. Hochstein a, Eric J. Olson b
a
b

3M Separation and Purification Sciences Division, 3M Center 236-1C-14, St. Paul, MN 55144-1000, United States
3M Corporate Research Analytical Laboratory, 3M Center 201-BS-03, St. Paul, MN 55144-1000, United States

a r t i c l e

i n f o

Article history:
Received 8 April 2021
Revised 12 July 2021
Accepted 22 July 2021
Available online 30 July 2021
Keywords:
Single-use chromatography
Anion exchange (AEX) chromatography
Polishing chromatography
Monoclonal antibodies (MAbs)
Downstream processing

Viral clearance

a b s t r a c t
Three efficiency tests for single-use AEX chromatography devices have been developed and applied to
six capsule formats of a new, salt tolerant, single-use AEX product. All the tests have been designed
to be performed with simple equipment and common reagents. By performing each of the three tests
on undamaged capsules and capsules intentionally damaged with small defects, in tandem with Phi-X174
challenges in a high-salt buffer, relationships between test results and viral clearance have been obtained.
A pre-use pressure-based installation verification test is simply performed during equilibration of the
device and effective at identifying gross bypass defects, for example, due to internal seal breakage. Passing
outcomes of a post-use installation validation bubble point test are associated with ≥ 5 log reduction
value (LRV) of viral clearance. A new, non-destructive, pre-use AEX capacity test involves challenging the
device with chloride ions and is orthogonal to the other two tests in that it can detect chemical defects,
as well as mechanical ones. Passing outcomes of this test correspond to > 2 LRV viral clearance and
provide in situ assurance of the expected AEX dynamic capacity prior to use. Selection of a pair of preuse and post-use tests can provide robust risk reduction with respect to viral clearance by single-use AEX
devices in biopharmaceutical purifications.
© 2021 3M Company. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
( />
1. Introduction
Flow-through anion exchange (AEX) chromatography is frequently used in biopharmaceutical purification processes for reduction of net-negatively charged host cell proteins (HCPs) and viral reduction as part of a validated viral clearance strategy [1,2].
AEX column chromatography is the technology most often used for
electrostatic viral clearance, particularly in commercial scale biopharmaceutical manufacturing, where columns have established a
long history of reliable and well understood performance [3]. Still,
validation of HCP and viral clearance by AEX columns in biopharmaceutical processes involves complexities which contribute significantly to operational and regulatory costs. Manufacturers must
be concerned with the possibility of micro-channeling in columns
which may result from defects in column packing, a concern routinely mitigated by in situ measurement of the asymmetry of the
elution peak resulting from the upstream pulse injection of an analyte and quantification of the height equivalent to a theoretical

∗

Corresponding author.
E-mail address: (J.F. Hester).

plate (HETP) derived from the elution peak retention time and
breadth [4–6]. Another concern is the potential for loss of viral
clearance with resin re-use, which may extend over hundreds of
use cycles with intervening cleaning procedures that have the potential to cause resin degradation [6–8]. An assessment of the effect of resin re-use on viral clearance is thus generally recommended on a product-by-product basis [1,6,7].
In recent years, the introduction of single-use AEX technologies has illuminated the potential for reduced regulatory and operational costs associated with flow-through AEX chromatography
[5,9–11]. Physically resembling and operated like filters, singleuse AEX products benefit from improved specific capacity and enhanced flow rates compared with columns due to the replacement
of diffusive kinetics with convective flow. These features have led
researchers to note the potential for simpler operation, decreased
processing times, and reduced buffer consumption leading to improved economics relative to columns. Additionally, their singleuse nature obviates validation costs associated with cleaning and
performance over repeat use cycles, including viral clearance performance. While single-use AEX products were initially used primarily in laboratory and process development activities, recent ad-

/>0021-9673/© 2021 3M Company. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
( />

J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

vances in ligand and media design have resulted in devices with
high specific capacity and robust performance across fluid conditions, making them viable alternatives to packed columns in commercial manufacturing [12].
With respect to the deployment of single-use AEX devices in
large-scale manufacturing, investigators have noted the need for
sensitive, in situ test methods capable of detecting device defects
that might result in reduced viral clearance. HETP and peak asymmetry studies used successfully on columns are not sensitive for
single-use devices, given their high flow rates, large mixing volumes, and short bed heights [13]. In defining test strategies for
single-use devices, it is useful to consider the viral clearance risks

that might occur. Viral clearance loss could occur due to mechanical bypass resulting from media or seal damage which might
originate in manufacturing or during shipping and handling, for
example. Alternatively, premature viral breakthrough could occur
in a mechanically integral capsule due to any of a few possible
chemical defects: regions of the media not properly chemically
functionalized during manufacturing or missing AEX media layers,
for example. Further, there is a time distribution of risk during
bioprocessing. To reduce the risk of processing valuable productcontaining fluid with a damaged capsule, resulting in a process
deviation and the need for costly re-processing, a non-destructive
pre-use efficiency test is desired. However, there is also a risk that
capsule damage could occur during processing, for example, due to
over-pressurization. To mitigate this risk, a post-use efficiency test
might be preferred.
A number of efforts have been made to develop and employ
in situ efficiency tests for single-use chromatography devices. Diffusion tests are commonly used, wherein the capsule media is
prewet with an aqueous solution, a constant upstream gas pressure is applied, and the gas diffusion rate across the media is
measured [14,15]. These are useful for detecting gross mechanical leaks such as seal damage or holes in the media. They would
fail to detect chemical defects, however. Multiple in situ tests have
been described that can characterize both mechanical and chemical integrity of single-use devices. Bind-and-elute type tests involve loading of the AEX device with an excess of a binding analyte, washing the media, and then eluting the analyte and characterizing the resulting elution profile [16,17]. Previously described
breakthrough-type tests involve challenging the device with a specific analyte while monitoring the concentration of the analyte in
the filtrate. Investigators have developed breakthrough-type tests
based on specialized analytes [18] as well as controlled pH changes
in the feed solution [19–23].
In situ efficiency tests developed to date generally have one or
more of the following challenges with respect to application at
commercial scale. Many of the tests, particularly those of the bindand-elute type, rely on subjective comparison of the analyte elution profile with that of a reference device. Quantitative determination of what deviation from the reference profile is indicative of a
defect expected to result in viral clearance loss, vs. inconsequential
media variability, is difficult, and such subjectivity is not generally
suitable to a Current Good Manufacturing Practices (cGMP) environment. Some of the tests utilize specialized reagents, not generally available in biopharmaceutical manufacturing plants, which
would need to be inventoried specifically for the tests. Finally, the

breakthrough-type tests rely on knowledge of the upstream fluid
volume between the fluid injection point and the downstream
breakthrough detector. This includes the fluid volume within the
AEX device as well as any upstream and downstream tubing. While
such holdup volume estimates are relatively straightforward and
commonly used on lab-scale equipment, they are problematic at
manufacturing scales, where long lengths of large-diameter tubing
may be utilized and where the tubing and capsule headspaces may
contain variable volumes of air bubbles, for example.

Herein are described three in situ efficiency tests for single-use
AEX devices from which may be selected an orthogonal set of preand post-use tests suitable for commercial manufacturing. Performance of the efficiency tests is assessed in the context of a recently commercialized, hybrid AEX device that combines a quaternary ammonium (Q) functional nonwoven AEX media with a novel
guanidinium (Gu) functional AEX membrane, achieving robust HCP
and virus reduction over wide ranges of fluid pH and ionic strength
[12]. The test methods have objective and quantitative pass/fail
criteria and utilize reagents commonly available in biopharmaceutical manufacturing environments. A non-destructive, pre-use
pressure-based efficiency test is conveniently performed during
equilibration of the device and is effective in identifying gross
mechanical defects (e.g., internal seal damage) that might result
in media bypass. A new breakthrough-type, pre-use AEX dynamic
capacity test [24] can detect fine mechanical or chemical defects, as well as provide in situ measurement of AEX capacity,
without the need to estimate holdup volume of the system. Finally, a post-use bubble point test identifies any mechanical defect
larger than the largest pore size of the AEX membrane pore size
distribution.
We present performance data for each of the three efficiency
tests on undamaged single-use devices of a variety of sizes, as well
as devices with purposefully introduced defects. To facilitate quantitative risk assessment, we present data relating efficiency test
outcomes with measured Phi-X174 viral clearance.
2. Experimental
2.1. Materials

2.1.1. Reagents
Sodium acetate (ACS, anhydrous), potassium chloride (ACS),
and Tris base (Biotechnology Grade) were purchased from VWR.
Sodium chloride (U.S.P.) was purchased from J. T. Baker. 1 N hydrochloric acid solution for buffer pH adjustment was purchased
from J. T. Baker. Filling solution for the chloride ion selective electrode (ISE) was purchased from ThermoFisher (Orion ionplus Optimum Results B, item no. 90 0 062). Filling solution for the potassium ISE was purchased from ThermoFisher (Orion ionplus Optimum Results E, item no. 90 0 065). All reagents were used as received. All salt solutions were prepared using deionized water provided by a Millipore Milli-Q water purification system which had a
resistivity > 18 M •cm.
2.1.2. Test capsules
3MTM Polisher ST single-use AEX capsules (3M Company, St.
Paul, MN) are available in six sizes (denoted BC1, BC4, BC25, BC170,
BC340, and BC1020, where the numeral in each designation refers
to the nominal media frontal surface area), as detailed elsewhere
[12]. The capsules contain a hybrid AEX media bed consisting of 4
layers of Q-functional nonwoven, having a combined bed depth of
0.35 cm, and 3 layers of Gu-functional membrane having a combined bed depth of 0.14 cm. Efficiency tests were performed on all
six capsule sizes.
Some pre-use AEX dynamic capacity tests were performed using a 2-piece threaded polycarbonate test housing that could restrain a variable number of 25 mm diameter media discs. The
housing comprised a top piece with a fluid inlet and a vent for
removing air from the housing and a bottom piece with a fluid
outlet. To simulate the media arrangement of a 3MTM Polisher ST
single-use capsule, 3 layers of Gu-functional membrane (FM, 3M
Company, St. Paul, MN) were placed in the downstream portion
of the housing. A plastic annular ring with an inner diameter of
19 mm and an outer diameter of 25 mm was placed on top of
2


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

pre-use AEX dynamic capacity test on all 3MTM Polisher ST capsule formats. A peristaltic pump (Cole-Parmer item no. ZM-0752220 with Masterflex L/S Easy-Load II Pump Head, item no. 7720060) supplied any of 1 M sodium acetate, 20 mM sodium acetate,
or 20 mM KCl solution to the capsule inlet, with valves positioned upstream of the pump as shown to select the challenge solution. For BC1, BC4, and BC25 capsules, potassium and chloride
ISE’s were mounted in a dual-probe plexiglass flow cell (Item No.
791101, FIAlab Instruments, Seattle, WA) with the potassium ISE
upstream of the chloride ISE. For larger BC170, BC340, and BC1020
capsules, the ISE’s were mounted in a purpose-built polycarbonate
flow cell with top and bottom pieces that were joined together and
sealed with an o-ring, defining a small internal volume in contact
with the ISE tips. Upstream and downstream pressure transducers
were monitored using a pressure monitor/transmitter (PMAT2A,
PendoTECH, Princeton, NJ). The ISE outputs were monitored using
a pH/conductivity monitor/transmitter (PDKT-ACCESS-PHCN, PendoTECH). Filtrate mass was monitored using a top-loading scale.
Mass, ISE output, and pressure data were automatically logged using PendoTECH PressureMAT data logging software (PMATP-GUI,
PendoTECH).
The capsule was initially filled with 1 M sodium acetate solution, with the vent open, until all air was removed from the capsule, after which the vent was closed. With reference to Fig. 1, in
a first step, bypass valve 6 was directed to bypass and the capsule
was flushed with 1 M sodium acetate to standardize the media by
replacing all bound counter-ions with acetate. In a second step, the
capsule was washed with 20 mM sodium acetate to remove from
the capsule housing the high acetate concentration from the previous step. In a third step, bypass valve 6 was set to direct the filtrate flow through the ISE flow cell and the flow rate of the 20 mM
sodium acetate feed solution was reduced to equilibrate the system
at the flow conditions of the subsequent challenge step. In a fourth
step, the capsule was challenged with a 20 mM potassium chloride
solution while the downstream ISE responses were monitored. Test
conditions for each capsule format are detailed in Table 1.
Potassium ions are unbound by the AEX media and pass
through the capsule with the challenge fluid front. Chloride ions
are bound by the media, displacing acetate ions. Fig. 2 is a plot of
exemplary resulting sets of potassium and chloride ISE responses
for three different 3MTM Polisher ST BC340 capsules. The opposing directionality of the K+ and Cl− breakthrough responses is explained by the Nernst equation, wherein the response of an electrochemical cell to an ionic analyte is proportional to the logarithm

of the analyte concentration multiplied by a term including the
charge of the analyte. Assuming collection of the filtrate during the
challenge step begins at the moment the feed solution is switched
from 20 mM sodium acetate solution to the KCl challenge solution,
the potassium ion breakthrough volume is the system holdup volume between the feed solution injection point and the ISE’s. The
chloride ion breakthrough volume is the volumetric throughput at
which the AEX capacity of the device is exhausted. The AEX capacity of the device is characterized by the net breakthrough volume
(Vnet ), the difference between the chloride and potassium breakthrough volumes. This value was normalized by the capsule surface area and expressed in terms of volume of filtrate per unit
area (1909 mL / 340 cm2 = 5.6 mL/cm2 for capsule 3 in Fig. 2)
or microequivalents of chloride per unit area based on the 20 mM
challenge solution concentration (112 μequiv/cm2 for capsule 3 in
Fig. 2). A “pass” criterion of ≥ 4 mL/cm2 (≥ 80 mequiv/cm2 ) was
selected on the basis of numerous trials conducted on capsules
containing Q-functional nonwoven and Gu-functional membrane
media spanning the media specification ranges for AEX capacity.
Thus, undamaged capsules containing media of both types at the
lower specification limit for AEX capacity are expected to pass the
above criteria.

those. Next, 4 layers of Q-functional nonwoven (FNW, 3M Company, St. Paul, MN) were placed on top of the ring. A second plastic annular ring was placed on top of the 4 layers of FNW. Finally,
an o-ring was placed on top of the media stack and the top portion of the housing was screwed down to restrain the media stack.
The resulting capsule had an effective frontal media surface area of
2.84 cm2 .
2.2. Phi-X174 preparation, filtration challenges, and enumeration
To prepare viral challenge solutions, a concentrated, filtersterilized Phi-X174 virus stock containing 1 × 1010 plaqueforming units (PFU)/mL was spiked to a target concentration of
> 1 × 108 PFU/mL into 50 mM Tris-HCl buffer, pH 8.0, adusted
to a conductivity of 20 mS/cm with NaCl. After flushing capsules
with 50 L/m2 of Tris-HCl, pH 8.0, 20 mS/cm buffer at an areanormalized flow rate of 1 mL/(cm2 -min), the capsules were challenged with 100 L/m2 of the spiked virus solution at the same flow
rate.
The input virus challenge solution and the filtered outputs were

enumerated using a plaque assay. Samples were serially diluted to
a concentration at which countable virus plaques (zones of clearing in a bacterial lawn caused by viral lysis) could be visualized
on a 100 mm agar plate (approximately 20–200 plaques/100 μl).
100 μL of the serially diluted virus sample and 50 μl of E. coli
13706 (ATCC) overnight host culture were mixed with 2.5 mL of
molten top agar (nutrient broth with 0.5% NaCl and 0.9% agar) and
poured on top of a 100 mm nutrient agar plate. Plates were incubated at 37 °C for 3-4 h and plaques were counted. The LRV was
calculated by taking the log of the input virus concentration minus
the log of the output concentration. Viral clearance results measured by this assay were expected to be accurate within approximately ±0.5 LRV [25].
2.3. Pre-use pressure-based installation verification test
For BC170, BC340, and BC1020 capsules, a non-destructive, preuse, pressure-based installation verification test was performed
as specified by the manufacturer [26]. Briefly, the pressure drop
across the capsule was monitored as it was flushed with 20 mM
sodium acetate at an area-normalized flow rate of 600 L/(m2 -h).
A pressure drop greater than or equal to 3 psid was considered
a “pass,” while a pressure drop less than 3 psid was considered
a “fail” possibly indicating a bypass within the capsule. To facilitate higher-rate testing for BC1, BC4, and BC25 capsules, the pressure drop was monitored as the capsule was flushed with 20 mM
sodium acetate at an area-normalized flow rate of 1,800 L/(m2 -h).
The measured pressure drop was then divided by 3 and the resulting value was compared with the ≥ 3 psid test criterion. (Pressure
drop was found to be linear with flux within this range of flow
rates.)
2.4. Pre-use AEX dynamic capacity test
2.4.1. Pre-use AEX dynamic capacity tests on capsules of all sizes
using a PendoTECH system
A non-destructive, pre-use efficiency test was conducted on
capsules by standardizing the AEX media with bound acetate
counter-ions and then challenging the capsules with dilute potassium chloride [24]. While, in the case of an undamaged singleuse capsule, this test fundamentally measures the charge density of the AEX media within, it is termed a “dynamic capacity” test because it is advantageously performed under flow conditions identical to those used when the capsule performs viral
clearance in use as recommended by the manufacturer. Fig. 1 is
a schematic of a tabletop setup that was used to perform this
3

J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Fig. 1. Schematic diagram of setup used to perform pre-use AEX capacity tests on all capsule formats. Solid lines denote fluid (tubing) connections, while dashed lines
denote electrical connections.

Fig. 2. Exemplary pre-use AEX capacity test data for three 3MTM Polisher ST BC340 (340 cm2 ) capsules [24]. Open symbols are potassium ISE responses (left axis) while
closed symbols are chloride ISE responses (right axis).

4


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Table 1
Steps and settings for performing the pre-use integrity test on any capsule size using a PendoTECH control and
data acquisition system. Numerals in black circles reference the corresponding labels in Fig. 1.
Step No.

Setting

Capsule Type
BC1

Setup

L/S® tubing size

14

ISE flow cell

BC4

BC25

BC170

BC340

BC1020

14

14

25

25

25

Small scale

Capsule filling

Flow rate, mL/min

1. Load

Feed fluid

1 M sodium acetate

Flow path

Valve ❻ to bypass

2. Wash

3

6

Pilot scale
20

500

4

12

80

500

1000

1000

Volume, mL

4

12

80

300

500

1000

Feed fluid

20 mM sodium acetate
Valve ❻ to bypass

Flow rate, mL/min

3

10

60

Volume, mL

15

45

60

Feed fluid

Flow rate, mL/min
Volume, mL

1000

1700

1700

4000

Valve ❻ to ISE flow cell
3

25

170

340

1000

≥3

≥ 10

≥ 50

≥ 500

≥ 500

≥ 1000

20 mM potassium chloride
Valve ❻ to ISE flow cell

Flow path

End criteria

1000

1

Feed fluid

Flow rate, mL/min

500

20 mM sodium acetate

Flow path

4. Challenge

500

Flow rate, mL/min

Flow path

3. Equilibrate

500

1

3

25

170

340

1000

End when significant Cl− ISE breakthrough response is observed

2.4.2. Pre-use AEX capacity tests on small capsules utilizing a Cytiva
ÄKTA chromatography system
Pre-use AEX dynamic capacity tests on media configurations
in the polycarbonate test housing were performed using a Cytiva
ÄKTA avant 25 chromatography system (Item No. 28930842, Cytiva). Potassium and chloride ISE’s were mounted in a dual-probe
plexiglass flow cell (Item No. 791101, FIAlab Intsruments, Seattle,
WA) with the potassium ISE upstream of the chloride ISE. The resulting ISE probe assembly was mounted on the ÄKTA chromatography system with HPLC tubing from the “Out1” position of the
ÄKTA outlet valve directed to the inlet of the side of the flow cell
containing the potassium ISE and tubing from the opposite side
of the flow cell directed to a waste container. The ÄKTA system
was outfitted with a Cytiva I/O-box E9 (Item No. 29011361, Cytiva)
to enable automatic data logging of the ISE outputs. Input to the
I/O-box E9 was provided by two analog cables with female jacks
(Item No. 290-1009-ND, Digi-Key). The ISE’s were connected to the
input cables using two BNC female interconnect jacks (Item No.
ARF1069-ND, Digi-Key). Automated data collection from the ÄKTA
system was provided by Cytiva UNICORN software, which was configured to log I/O-box E9 inputs according to the manufacturer’s instructions. Deionized water was setup on system pump “A” at position “A1.” A 0.5 M sodium acetate equilibration solution was setup
on system pump “B” at position “B1.” The 20 mM KCl challenge solution was setup on the sample pump at position “S1.” Generally,
Cytiva ÄKTA avant and ÄKTA pure chromatography system models
are suitable for performing the pre-use AEX dynamic capacity test
on 3MTM Polisher ST BC1, BC4, and BC25 capsules.
The polycarbonate housing was assembled as described in
Section 2.1.2 and the resulting filter assembly was mounted on
a column position of the ÄKTA system. With the vent open, 0.5
M sodium acetate solution from inlet position “B1” was pumped
slowly into the housing until all the air was removed and then

the vent was closed. The media was then equilibrated by pumping
24 mL of 0.5 M sodium acetate through the housing at a flow rate
of 12 mL/min with the filtrate directed to outlet valve position “W”
(waste). The ÄKTA system was then configured to deliver a 20 mM
sodium acetate wash solution as a gradient comprising 4% of 0.5
M sodium acetate solution from inlet position “B1” and 96% deionized water from inlet position “A1.” 12 mL of the wash solution
was pumped through the housing at a flow rate of 6 mL/min with
the filtrate directed to outlet valve position “W.” Then, the outlet
valve position was switched to “Out1,” to direct the filtrate to the
ISE assembly, and 12 mL of the 20 mM sodium acetate wash solution was pumped through the housing and ISE assembly at a flow
rate of 3 mL/min. Finally, the injection valve was moved to change
the inlet flow source to the 20 mM KCl challenge solution at sample pump position “S1,” and the challenge solution was pumped
through the capsule at 3 mL/min. The potassium and chloride ISE
responses were monitored, and flow was continued until complete breakthrough responses had been detected by both probes.
For each run, the filtrate volume at the initial inflection point of
each of the potassium and chloride breakthrough responses was
recorded as the corresponding breakthrough volume, and the AEX
capacity was calculated as described in Section 2.4.1.
2.5. Post-use installation validation bubble point test
A post-use installation validation bubble point test was performed on capsules as specified by the manufacturer [26]. Briefly,
the capsule headspace was gravity drained and then the inlet was
connected to a Sartocheck® 4 Plus Filter Tester (Sartorius) while
a length of tubing connected to the capsule outlet was immersed
in water. An upstream air pressure of 50 mbar was applied, and
the air pressure was ramped up in 50 mbar increments until the
bubble point pressure of the capsule was detected by visual obser5


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

vation of a stream of bubbles emerging from the outlet tubing. A
bubble point pressure greater than or equal to 900 mbar was considered a “pass,” while a bubble point pressure less than 900 mbar
was considered a “fail” possibly indicating a mechanically defective capsule. Note that, while some single-use AEX device manufacturers recommend similar gas diffusion type tests as pre-use
tests [14,15], the bubble point test described above must be considered destructive for 3MTM Polisher ST and thus cannot be performed as a pre-use test, as it can introduce air between the FNW
layers that may not be reliably removed.

relatively large capsule headspaces to enable rapid flow and handling of turbid fluids [12]. The large resulting mixing volume upstream of the AEX media is responsible for the broad breakthrough
curves relative to columns. Selection of acetate as the standardizing anion is important; KCl challenges conducted on capsules standardized with sodium bicarbonate or sodium sulfate featured earlier and broader chloride breakthrough curves and reduced distinction between integral and damaged capsules (data not shown).
This might be expected, since acetate has a lower selectivity coefficient with respect to the Q-functional media than chloride, resulting in a self-sharpening ion exchange boundary within the media,
whereas more tightly-binding standardizing anions than chloride
would be expected to result in an ion exchange boundary spreading with progression through the media [28].
Attempts were made to calibrate the ISE’s such that the breakthrough curves could be plotted in units of ion concentration
rather than as raw ISE electrical responses as in Fig. 1. These attempts were frustrated by significant ISE response drift over time
originating from a number of factors. Among these, both ISE’s are
essentially measuring “zero” concentration prior to breakthrough
of their respective analyte ions, the resulting cell potential of
which is not well defined. This accounts for the variable vertical positions of the breakthrough curves in Fig. 1 and other figures herein. A second factor is relatively rapid drifts in the quantitative ISE responses over subsequent experiments, for example,
due to plasticization of the polymeric membrane of the potassium
ISE by acetate [29]. Due to these challenges, ISE responses herein
are plotted as raw electrical responses, since the primary response
attributes of value are the filtrate volumes at which K+ and Cl−
breakthrough occur.

2.6. Controlled damage of AEX capsules
To create various controlled levels of media damage, some
3MTM Polisher ST BC1, BC4, BC25, BC170, BC340, and BC1020 capsules were prepared with small, controlled defects by introducing holes through the full media stack using blunt-tipped needles (Small Hub RN Needle, Point Style 3, Hamilton, Reno, NV) of
various diameter ranging from 32 gauge (0.11 mm) to 18 gauge

(0.84 mm). In the case of BC1, BC4, and BC25 laboratory capsules,
the holes were introduced in the stacked media prior to capsule
assembly. In the case of BC170 capsules, holes were introduced in
the media in the internal lenticular filter element prior to assembling the capsule. In the case of BC340 and BC1020 capsules, holes
were introduced through one entire lenticular element, including
media stacks on both sides of the element, prior to assembling the
capsule. One BC170 capsule was assembled using an undamaged
lenticular filter element but with a purposeful misalignment of an
o-ring that forms a seal between the filter element and the housing; this was expected to create a gross mechanical bypass within
the capsule.

3.2. Detection of missing media layers using pre-use AEX dynamic
capacity test

3. Results and discussion
Fig. 3 is a plot of potassium and chloride ISE responses during
the AEX dynamic capacity test for a complete media stack of 4 layers of FNW and 3 layers of FM in a 25 mm test housing (3 replicates), followed by ISE responses for sequential trials in which one
layer of media was removed from the housing in each trial. The
ISE response inflection volumes were used to compute volumetric
capacities as detailed in Table 3.
The full media stack exhibits a reproducible and passing
(≥4 mL/cm2 ) volumetric chloride challenge capacity. Media stacks
with “missing” layers exhibit capacities that fail the test criterion and steadily decrease as media layers are removed. While
the chloride breakthrough volume shifts significantly with the removal of each media layer, the potassium breakthrough volume
shifts only slightly, reflecting the smaller hold-up volume as the
housing volume decreases with layer removal. An empty housing
exhibits nearly coincident potassium and chloride breakthrough
events. These results illustrate the capability of the non-destructive
AEX dynamic capacity test to detect missing functional media layers and characterize the AEX capacity of single-use capsules prior
to use, and also the hold-up volume independence of the method.

The potassium ISE response curves exhibit a second, minor inflection coincident with chloride breakthrough. This is a result of
the change in chloride concentration at the reference electrode liquid junction of the potassium ISE, the internal filling solution of
which is aqueous sodium chloride [30].

3.1. Performance and repeatability of the pre-use AEX dynamic
capacity test on integral capsules
Five sequential AEX dynamic capacity tests were performed on
each of three 3MTM Polisher ST BC340 capsules, produced using
the same AEX media lots, as described in Section 2.4.1. One representative set of potassium and chloride responses for each capsule appears in Fig. 1. Potassium ISE responses featured a gradual, monotonically diminishing value prior to arrival of potassium
ions at the ISE flow cell, followed by the rapid onset of an increasing ISE response which was recorded as the K+ breakthrough volume as illustrated in Fig. 1. The internal BC340 capsule volume
of 730 mL [27] defines an expected lower limit of the K+ breakthrough volume. In practice, measured potassium breakthrough
volumes were somewhat larger and variable due to the volume
of tubing between the challenge solution injection point and the
ISE flow cell and also due to the fact that the test operator collected variable volumes of filtrate on the mass balance during the
20 mM sodium acetate equilibration step before switching the feed
to the KCl challenge solution. The chloride ISE responses featured
a gradual, monotonically increasing value prior to chloride breakthrough, followed by the onset of a monotonically decreasing response which was recorded as the Cl− breakthrough volume as illustrated in Fig. 1. K+ and Cl− breakthrough volumes and calculated dynamic capacities for all 15 trials are detailed in Table 2.
The test was reproducible with a coefficient of variation well below 10 percent.
3MTM Polisher ST capsules are designed for single-use AEX
chromatography in a flow-through mode. They feature shallow AEX
bed depths relative to columns, and engineering trade-offs have
been made to facilitate additional performance features, such as

3.3. Detection of damaged media using pre-use AEX dynamic
capacity test
Fig. 4 is a plot of potassium and chloride ISE responses during the AEX dynamic capacity test for an undamaged BC1 capsule and for BC1 capsules containing damaged media stacks, each
6


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Table 2
ISE breakthrough volumes and calculated AEX dynamic binding capacities for 15 trials conducted on three 3MTM Polisher ST BC340 capsules produced
using the same AEX media lots.
Capsule
1

Capsule 1
Capsule 1
Capsule 1
2

Capsule 2
Capsule 2
Capsule 2
3

Capsule 3
Capsule 3
Capsule 3

Trial

K+ Breakthrough Volume, mL

1
794
2

859
3
868
4
756
5
837
Mean
Standard Deviation
Coefficient of Variation
1
848
2
908
3
1048
4
922
5
824
Mean
Standard Deviation
Coefficient of Variation
1
811
2
776
3
1917
4

786
5
902
Mean
Standard Deviation
Coefficient of Variation

Cl− Breakthrough Volume, mL

Volumetric capacity, mL/cm2

Cl− capacity, μequiv/cm2

2470
2597
2744
2433
2768

4.9
5.1
5.5
4.9
5.7
5.2
0.3

99
102
110

99
114
105
7

2802
2799
2956
2875
2705

5.7
5.5
5.6
5.7
5.5
5.6
0.1

2720
2711
3702
2632
2545

5.6
5.7
5.2
5.4
4.8

5.4
0.3

6.6%
115
110
112
115
111
113
2
2.0%
112
114
105
109
97
107
7
6.4%

Fig. 3. Plot of chloride ISE response curves (top set of curves) and potassium ISE response curves (bottom set of curves) for 25 mm test housing containing various sets of
AEX functional media layers. Data collected using an ÄKTA avant 25 chromatography system.

7


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Table 3
ISE breakthrough volumes and calculated AEX dynamic binding capacities for 25 mm test housing containing various sets of AEX functional media layers.
Media Construction
4 FNW + 3 FM

3 FNW + 3 FM
2 FNW + 3 FM
1 FNW + 3 FM
3 FM
2 FM
1 FM
Empty housing

Trial 1
Trial 2
Trial 3

K+ Breakthrough Volume, mL

Cl- Breakthrough Volume, mL

Volumetric capacity, mL/cm2

Cl- capacity, μequiv/cm2

3.33
3.59
3.53
3.26

3.18
3.14
2.71
2.67
2.63
2.37

16.07
16.37
16.41
14.16
12.75
10.47
7.68
6.42
4.67
2.46

4.5
4.5
4.5
3.8
3.4
2.6
1.8
1.3
0.7
0.0

90

90
91
77
67
52
35
26
14
1

Fig. 4. Plot of chloride ISE response curves (top set of curves) and potassium ISE response curves (bottom set of curves) for BC1 (1 cm2 ) capsules containing media stacks
pierced with blunt-tipped needles of various diameter. All curves have been shifted on the x-axis such that the potassium breakthrough inflection point was set at zero
volume. Response curves have been intentionally vertically offset for easy comparison. Y-axis tick marks are at intervals of 0.1 mA. Hole sizes and net breakthrough volumes
for each curve are indicated. Data collected using the PendoTECH setup.

of which was pierced once with a blunt-tipped needle varying in
diameter from 32 gauge (0.11 mm) to 22 gauge (0.41 mm). The
media stack comprises two AEX media with different characteristics with respect to these piercing defects. Whereas piercing of the
FM creates a defined, roughly circular hole, the FNW swells upon
wetting in buffer and appears capable of “healing” small puncture
defects upon swelling. At the largest hole size, it appears that a
relatively well-defined low-pressure path is created in the media
stack, and chloride breakthrough is observed to be nearly coincident with potassium breakthrough. At smaller hole sizes, chloride breakthrough occurs earlier than for an undamaged capsule
and the breakthrough curves broaden. This is consistent with the

superposition of fast chloride breakthrough within a low-pressure
path comprising partially “healed” FNW layers and typical chloride
breakthrough within the remainder of the media. The pre-use AEX
dynamic capacity test is quite sensitive to small defects, with a defect as small as 0.003% of the frontal media area having a volumetric capacity <4 mL/cm2 and failing the test criterion.
3.4. Relation of pre-use pressure-based test, pre-use AEX dynamic

capacity test, and post-use bubble point test results with viral
clearance
A total of 111 capsules of various format were made, one of
which contained a purposefully misaligned internal seal and many
8


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Fig. 5. Summary plot of observed pre-use pressure-based installation verification test values vs. measured Phi-X174 viral clearance. Upward arrows denote viral clearance
experiments in which no virus plaques were observed after plating of the filtrate, and the minimal viral clearance was thus defined by the measured concentration of the
challenge. Dashed line highlights the pre-use installation verification test criterion.

Table 4
Undamaged and damaged capsules tested in experiments relating efficiency test outcomes to viral
clearance.
Needle Size
Gauge

Diameter, mm

Undamaged
32
30
28
26.5
25
22

18

0.108
0.159
0.184
0.235
0.260
0.413
0.838

Total Capsules Tested

No. of Media Lot Combos / No. of Replicates per Lot Combo
BC1

BC4

BC25

BC170

3
3
3
3
3
3
3
3

2
2
2
2
2
2
2
2

1
1
1
1
1
1
1
1

1
1
1
1
1

/
/
/
/
/
/

/
/

2
2
2
2
2
2
2
2

48

/
/
/
/
/
/
/
/

2
2
2
2
2
2
2

2

32

of which were intentionally damaged with blunt-tipped needles
as detailed in Table 4, to examine the relationship between each
of the three efficiency tests, performed on integral and defective
capsules, with Phi-X174 viral clearance in a high conductivity challenge. An effort was made to ensure the capsule population in this
study was representative of the “real world” variability expected
to exist among manufactured capsules. For example, the FM media primarily responsible for AEX viral clearance in these devices,
particularly at high conductivity [12], is released in manufacturing
on the basis of its measured bovine serum albumin (BSA) dynamic
binding capacity (DBC), which has minimum and maximum release
values of 7.9 and 12.0 mg/cm2 , respectively [27]. Each of the 111
capsules in this study was built using one of 11 different lots of

/
/
/
/
/
/
/
/

2
2
2
2
2

2
2
2

16

/
/
/
/
/
1/
1/
7

1
1
1
1
1
1
1

BC340

BC1020

1/1
1/1
1/1

1/1

1/1
1/1
1/1

4

3

FM with BSA DBC spanning nearly the entire allowable range of
BSA DBC, from 8.2 to 11.9 mg/cm2 .
Each capsule was tested according to the following workflow. First, a pre-use pressure-based installation verification test
was conducted using 20 mM sodium acetate during the preconditioning flush required by the manufacturer to remove glycerin used to stabilize the functional media. Second, a pre-use AEX
dynamic capacity test was performed using the PendoTECH setup.
Third, a Phi-X174 challenge was conducted under high-salt buffer
conditions (50 mM Tris-HCl, pH 8.0, adjusted to 20 mS/cm with
NaCl). Finally, a post-use installation validation bubble point test
was conducted.

9


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Fig. 6. Summary plot of observed pre-use AEX dynamic capacity test values vs. measured Phi-X174 viral clearance. Upward arrows denote viral clearance experiments in
which no virus plaques were observed after plating of the filtrate, and the minimal viral clearance was thus defined by the measured concentration of the challenge. Vertical
dashed line highlights the pre-use AEX dynamic capacity test criterion. Horizontal dashed line highlights a 2-LRV viral clearance level.

Figs. 5–7 are summary plots of pre-use pressure-based installation verification test, pre-use AEX dynamic capacity test, and postuse installation validation bubble point test values, respectively,
versus Phi-X174 LRV for undamaged and needle-damaged capsules.
The pre-use pressure-based installation validation test successfully
detected the gross bypass resulting from the misaligned seal in the
BC170 capsule damaged in that way, providing a failing pressure
drop of 0.6 psid as compared to a passing pressure drop of 12.9
psid for an undamaged and properly assembled BC170 capsule.
However, the data in Fig. 5 indicate this simple pressure-based test
is not capable of detecting small defects (like fine needle piercings)
that can result in substantial loss of viral clearance.
As shown in Fig. 6, the pre-use AEX dynamic capacity test is
more effective at detecting small defects, providing a positive correlation between measured AEX dynamic capacity and viral clearance. The correlation is not excellent (linear R2 = 0.67), which
might be expected based on the manufacturing variability of the
AEX media as noted above, the maximum measurable viral clearance of just greater than 7 LRV defined by the challenge concentration, and the expected viral clearance plaque assay accuracy
of about ±0.5 LRV. Still, based on the testing of 110 capsules of
all six formats selected to be representative of approximately the
full range of manufacturing variability, passage of the test criterion was associated with greater than 2 LRV viral clearance in a
high-salt buffer. Additionally, this test is capable of detecting not
only mechanical defects, but also chemical ones, and provides a
non-destructive pre-use measurement of the AEX capacity of the
single-use capsule.
As shown in Fig. 7, the post-use installation validation bubble
point test is the most effective test for characterizing small me-

chanical defects. In one case, a BC170 capsule with a 28-gauge hole
had a post-use bubble point pressure greater than the test criterion and had a measured viral clearance of > 4.84 LRV. No virus
plaques were observed in the plated viral challenge filtrate, however, and the value of 4.84 LRV was thus a lower limit of the viral
clearance because the concentration of the spike in this particular challenge was insufficient to measure > 5 LRV. In all other of
the 110 capsules tested, passage of the post-use bubble point test

criterion was associated with ≥ 5 LRV Phi-X174 clearance.
4. Conclusion
Three efficiency tests have been developed and applied to characterize mechanical and chemical defects in six capsule formats of
3MTM Polisher ST, a single-use AEX chromatography product. Two
non-destructive, pre-use tests can be applied to reduce the risk of
processing product-containing fluid with a damaged or defective
capsule. A pre-use pressure-based installation verification test is
conveniently applied, with minimal equipment (only a pump and
a pressure gauge or transducer), during the pre-conditioning flush
procedure required to remove glycerin stabilizer from the capsule
media and equilibrate the capsule for use. This simple test is capable of detecting gross mechanical defects, such as broken internal
seals, that might result from shipping damage, for example.
A new pre-use AEX dynamic capacity test is somewhat more
complicated to execute, though it uses common reagents. This
test is more effective at reducing the risk of processing productcontaining fluid using a capsule containing even very small defects,
with passing test results associated with > 2 LRV viral clearance
in a high-salt buffer challenge. In addition, the AEX dynamic ca10


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

Fig. 7. Summary plot of observed post-use installation validation bubble point test values vs. measured Phi-X174 viral clearance. Upward arrows denote viral clearance
experiments in which no virus plaques were observed after plating of the filtrate, and the minimal viral clearance was thus defined by the measured concentration of the
challenge. Vertical dashed line highlights the post-use bubble point test criterion. Horizontal dashed line highlights a 5-LRV viral clearance level.

pacity test can detect chemical defects as well as mechanical ones
and provides an in situ measurement of the capsule’s AEX capacity
prior to running product-containing fluid.

A post-use installation validation bubble point test provides the
greatest risk reduction with respect to mechanical defects, a passing test result corresponding to ≥ 5 LRV viral clearance in a highsalt buffer challenge. The bubble point test must be considered destructive for single-use devices like 3MTM Polisher ST that contain
multiple layers of functional media. Running it after capsule use,
however, provides risk reduction with respect to any capsule damage that may have occurred either before or during use.
From the above three tests may be chosen a pair of pre-use
and post-use efficiency tests that can be performed to substantially reduce the risk of viral clearance loss due to defects or damage in single-use AEX devices. For operators concerned with minimizing risks associated with potential use of a single-use capsule
to process product-containing fluid, an orthogonal combination of
the pre-use AEX dynamic capacity test and the post-use bubble
point test provides strong reduction of risk pre-use, as well as in
situ characterization of device capacity. Alternatively, a combination of the pre-use pressure-based installation verification test and
the post-use bubble point test is very simple to perform and provides effective pre-use detection of the types of capsule defects expected to be the most frequent (e.g., gross bypass due to internal
seal breakage resulting from shipping damage). Both combinations
of tests provide an ultimate level of risk reduction associated with
at least 5 LRV viral clearance in a high-salt buffer challenge according to a data set based on 110 capsules tested in this study.
In either case, it is recommended that a fluid reprocessing option
be included in regulatory filings for any AEX step, in case post-use
validation were to detect the presence of a capsule defect.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

CRediT authorship contribution statement
Jonathan F. Hester: Conceptualization, Methodology, Software,
Validation, Formal analysis, Investigation, Writing – original draft,
Writing – review & editing, Visualization. Xinran Lu: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation. Jacob D. Calhoun: Methodology, Validation, Formal analysis, Investigation, Data curation. Rebecca A.
Hochstein: Methodology, Validation, Formal analysis, Investigation,
Writing – original draft. Eric J. Olson: Methodology, Writing – review & editing.
References
[1] , Viral safety evaluation of biotechnology products derived from cell lines of

human or animal origin, in: Proceedings of the International Conference on
Harmonisation of Technical Requirements for Registration of Pharmaceuticals
for Human Use: Geneva, 1999 Switzerland.
[2] Center for Biologics Evaluation and Research. Points to consider in the
manufacture and testing of monoclonal antibody products for human Use.
U.S. Food and Drug Administration: Rockville, MD, 1997; .
gov/regulatory-information/search-fda-guidance-documents/points-considermanufacture- and- testing- monoclonal- antibody- products- human- use.
[3] H.F. Liu, J. Ma, C. Winter, R. Bayer, Recovery and purification process development for monoclonal antibody production, MAbs 2 (5) (2010) 480–499.
[4] JC. Giddings, Dynamics of Chromatography, Part 1: Principles and Theory, Marcel Dekker, Inc., 1965.
11


J.F. Hester, X. Lu, J.D. Calhoun et al.

Journal of Chromatography A 1654 (2021) 462445

[5] G.R. Miesegaes, S.C. Lute, E.K. Read, KA. Brorson, Viral clearance by
flow-through mode ion exchange columns and membrane adsorbers, Biotechnol. Prog. 30 (1) (2014) 124–131.
[6] L. Norling, S. Lute, R. Emery, W. Khuu, M. Voisard, Y. Xu, Q. Chen, G. Blank,
K. Brorson, Impact of multiple re-use of anion-exchange chromatography media on virus removal, J. Chromatogr. A 1069 (1) (2005) 79–89.
[7] R.M. O’Leary, D. Feuerhelm, D. Peers, Y. Xu, GS. Blank, Determining the useful
life of chromatography resins: prospective small-scale studies, BioPharm 14 (9)
(2001) 10–18.
[8] B.D. Kelley, J. Jakubik, S. Vicik, Viral clearance studies on new and used chromatography resins: critical review of a large dataset, Biologicals 36 (2) (2008)
88–98.
[9] V. Orr, L. Zhong, M. Moo-Young, C.P. Chou, Recent advances in bioprocessing application of membrane chromatography, Biotechnol. Adv. 31 (4) (2013)
450–465.
[10] T. Vicente, M.F.Q. Sousa, C. Peixoto, J.P.B. Mota, P.M. Alves, M.J.T. Carrondo, Anion-exchange membrane chromatography for purification of rotavirus-like particles, J. Membr. Sci. 311 (1-2) (2008) 270–283.
[11] J. Weaver, S.M. Husson, L. Murphy, S.R. Wickramasinghe, Anion exchange
membrane adsorbers for flow-through polishing steps: part I. Clearance of

minute virus of mice, Biotechnol. Bioeng. 110 (2) (2013) 491–499.
[12] J. Hester, J. Heitkamp, M. Peters, Z. Jokondo, J. Rasmussen, Streamlined polishing and viral clearance using a new hybrid, biomimetic single-use anion
exchanger, BioProcess Int. 18 (10) (2020) 70–76.
[13] J.X. Zhou, F. Solamo, T. Hong, M. Shearer, T. Tressel, Viral clearance using disposable systems in monoclonal antibody commercial downstream processing,
Biotechnol. Bioeng. 100 (3) (2008) 488–496.
[14] Sartorius Stedim Biotech GmbH. Operating instructions: Sartobind® Q and S
Membrane Adsorbers, Sartorius order number 85037-549-25, 2015.
[15] Pall CorporationMustang® Q, S and E Single-Use Capsules: Assembly and
Installation Instructions, Pall Corporation Document Number USD, 2021 2933c.
library/non- gated/
instructions- for- use/19- 10212- USD2933- c- Mustang- Q- S- E- Capsules- IFU- EN.
pdf?_ga=2.126645837.1598829374.1600449057-86485960.1600449057.
[16] S. Nochumson, P. Levison, inventors; Pall Corporation, assignee. System and
method for testing chromatography media and devices. U.S. Patent 7,732,216.
June 8, 2010.
[17] F.M. Plieva, I.N. Savina, S. Deraz, J. Andersson, I.Y. Galaev, B. Mattiason, Characterization of supermacroporous monolithic polyacrylamide based matrices
designed for chromatography of bioparticles, J. Chromatogr. B 807 (1) (2004)
129–137 .

[18] M.W. Phillips, inventor; Millipore Corporation, assignee. Method for determining an effective Peclet number for a membrane adsorber device. U.S. Patent
7,281,410. October 16, 2007.
[19] W. Demmer, R. Faber, H.H. Hoerl, A. Thiefes, inventors; Sartorius Stedim Biotech GmbH, assignee. Method for qualifying a non-particulate ionexchanger adsorber. U.S. Patent 8,969,094. March 3, 2015.
[20] A. Podgornik, J. Vidic, J. Jancar, N. Lendero, V. Frankovic, A. Strancar, Noninvasive methods for characterization of large-volume monolithic chromatography
columns, Chem. Eng. Technol. 28 (11) (2005) 1435–1441.
[21] N. Lendero, J. Vidic, P. Brne, A. Podgornik, A. Strancar, Simple method for determining the amount of ion-exchange groups on chromatographic supports, J.
Chromatogr. A 1065 (1) (2005) 29–38.
[22] N. Lendero, J. Vidic, P. Brne, V. Frankovic, A. Strancar, A. Podgornik, Characterization of ion exchange stationary phases via pH transition profiles, J. Chromatogr. A 1185 (1) (2008) 59–70.
[23] P.R. Gavara, R. Cabrera, R.R. Vennapusa, M. Grasselli, M. Fernandez-Lahore,
Preparation, characterization, and process performance of composite fibrous
adsorbents as cation exchangers for high throughput and high capacity bioseparations, J. Chromatogr. B 903 (2012) 14–22.

[24] J.F. Hester, E.J. Olson, F.E. Porbeni, A.W. Vail, A.M. Voloshin, inventors; 3M Innovative Properties Company, assignee. Method for testing a chromatography
devices used for ion exchange. WO Patent Application 2020/128797. June 25,
2020.
[25] H. Ruppach, Log10 reduction factors in viral clearance studies, BioProcess. J. 12
(4) (2014) 24–30.
[26] 3M Purification Inc. 2020. Installation qualification technical brief, 3MTM
polisher ST series capsules, 3M document number 70-0216-0019-7 INS
Rev 0920; polisher- stinstallation- qualificaiton- tech- brief.pdf.
[27] 3M Purification Inc. 2021. 3MTM Polisher ST Regulatory Support File, 3M document number 10-698362 REV 0721.
[28] F. Helfferich, Theories of ion-exchange column performance: a critical study,
Angew. Chem. Int. Ed. 1 (8) (1962) 440–453.
[29] U. Oesch, W. Simon, Life time of neutral carrier based ion-selective liquid-membrane electrodes, Anal. Chem. 52 (1980) 692–700.
[30] F. Bretschneider, J.R. de Weille, in: Introduction to Electrophysiological Methods and Instrumentation, Academic Press, 2006, pp. 103–131.

12