Versatile Methods to Dispense Submilligram Quantities of Solids Using Chemical-Coated Beads for High-Throughput Experimentation

Article abstract of DOI:10.1021/acs.oprd.9b00213

High-throughput experimentation is a technique for screening multiple reaction conditions in parallel at micro or nanoscale without depleting precious starting materials. However, assembling a comprehensive screening set often involves the distribution of large number of solid reagents with diverse physical properties in small quantities. Automated solid dispensing, especially at submilligram scale, has long been a challenge with no practical and reliable solutions. This paper describes the use of our newly developed chemical-coated beads technology to provide a universal approach to the solid handling problem. This technology, when combined with an automated solid dispensing platform or calibrated scoops, can dispense submilligram quantities of a variety of solids with efficiency and adequate accuracy.

Full text of DOI:10.1021/acs.oprd.9b00213

Subscriber access provided by UNIV OF SOUTHERN INDIANA
Full Paper
Versatile Methods to Dispense Sub-Milligram Quantities of Solids
using Chemical Coated Beads for High-Throughput Experimentation
M Cynthia Martin, Gashaw M Goshu, jeffery hartnell, Collin Morris, Ying Wang, and Noah P Tu
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.9b00213 • Publication Date (Web): 30 Jul 2019
Downloaded from pubs.acs.org on July 30, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Versatile Methods to Dispense Sub‐Milligram Quantities of
Solids using Chemical Coated Beads for High‐Throughput
Experimentation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
M. Cynthia Martin, Gashaw M. Goshu, Jeffery R. Hartnell, Collin D. Morris, Ying Wang and
Noah P. Tu*
*AbbVie Inc., 1 North Waukegan Road, North Chicago, Illinois 60064, United States
ABSTRACT: High‐throughput experimentation (HTE) is a technique for screening multiple
reaction conditions in parallel at micro‐ or nanoscale without depleting precious starting
material. However, assembling a comprehensive screening set often involves the
distribution of large number of solid reagents with diverse physical properties in small
quantities. Automated solid dispensing, especially at sub‐milligram scale, has long been a
challenge with no practical and reliable solutions. This paper describes the use of our newly
developed chemical coated beads (ChemBeads) technology to provide a universal approach
to the solid handling problem. This technology, when combined with an automated solid
dispensing platform, or calibrated scoops, can dispense sub‐milligram quantities of a
variety of solids with efficiency and adequate accuracy.
KEYWORDS: high throughput experimentation, chemical coated beads, sub‐milligram solid
dispensing, automation, microscale research, ChemBeads
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 22
1
2
3
4
INTRODUCTION
5
6
7
8
9
In early drug discovery, one of the primary goals is the development and optimization of
synthetic reaction conditions for rapid generation of a new single compound or a collection
of diverse compounds via parallel library synthesis.1 The traditional iterative process of
experimentation can be tedious, time‐consuming and costly for the following reasons: (1) it
requires time and manual labor to investigate a viable synthetic route to targeted
compounds and carry out the solid handling for a large array of small scale reactions, (2)
there is a limited supply of advanced intermediates needed for the synthesis of targeted
compounds, and, (3) there is a lack of automation for quick compilation of the screening
sets and to set up reactions at microscale.2 To remedy this situation, the art of high‐
throughput experimentation (HTE) is progressively being employed across the
pharmaceutical industry3 and academic laboratories.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Microscale HTE is a technique that aims at screening multiple reaction conditions in
parallel with minimal use of precious starting material (1 mg or less per reaction).5 Ideally,
a microscale HTE approach should survey reaction conditions broadly as a starting point
for chemists to further investigate and optimize the transformation. However, assembling a
comprehensive screening set often involves the distribution of large number of reagents in
small quantities. Previously the dispensing of materials, as stock solutions, was carried out
by a liquid handler followed by evaporation of solvent, rendering it time‐consuming, labor
intensive and limited to only soluble reagents.3,5 Automated solid dispensing, especially in
sub‐milligram scale, has long been a challenge with no practical and reliable solutions.
During our quest for a practical solution to the solid dispensing challenge, we realized
that it is extremely difficult for a single robotic platform to dispense a large variety of solid
ACS Paragon Plus Environment

Page 3 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
reagents with diverse physical properties. The difficulties compounded when the targeted
weight is in the low to sub‐milligram range, which is most desirable for setting up
microscale HTE. The Enabling Technologies Consortium High Throughput Experimentation
Working Group (HTE WG) described their collaborative studies on solid dispensing
technologies on multiple platforms using solids with diverse physical properties commonly
encountered in a pharmaceutical research and development laboratory.6 They reported that
while there were strengths and weaknesses of each robotic platform studied, there was a
major limitation for these systems to accurately dispense low target masses (2 mg or less).
To the best of our knowledge, there is no suitable technology that can effectively dispense
sub‐milligram quantities of solid reagents with various physical properties commonly used
in pharmaceutical and academic laboratories. This industry‐wide challenge prompted us to
tackle the root cause of the problem towards a universal solution. One approach is to unify
the diverse physical properties of solids into one uniform and favorable form, hence
simplifying the solid dispensing process to enable a single platform to dispense a variety of
solid reagents.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We recently reported a novel technology whereby solid reagents (guest particle) are
coated on glass beads (host particle) to enable dispensing of various solid reagents at sub‐
milligram scale for high throughput experimentation (ChemBeads).7 Herein we report the
extension of the technology to include polystyrene beads as the host material and detailed
studies of this technology in dispensing diverse solid reagents. ChemBeads technology
utilized a dry powder coating technique to improve powder flowability, content uniformity,
and dissolution properties of native solid reagents.8 During this dry powder coating
process, solid reagent adheres to the surface of the host material (glass or polystyrene
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 22
1
2
3
4
5
6
7
8
9
beads) via Van der Waals forces, resulting in ChemBeads that retained the high flowability
of the host material.9 In essence, the “transformed” physical properties and the low mass
loading (1‐20% loadings) of various solid reagents to that of the host materials can: (1)
render ChemBeads of different solid reagents to have similar physical properties and
enable the development of a single solid dispensing protocol for diverse solids in
ChemBeads form and, (2) deliver sub‐milligram amount of native reagents using milligram
target mass of ChemBeads. In addition, the low variations in densities of the glass and
polystyrene beads (regardless of the solid reagent coated) enabled the use of a volume to
“weigh” out ChemBeads. Using a suite of 3D printed scoops with calibrated scoop volume
(see Supporting Information Figure S7), ChemBeads can be dispensed manually without
the need of an analytical balance.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To further evaluate the effectiveness of the ChemBead approach, we set up experiments to
dispense the seven test solids as chosen by the HTE WG.
MATERIALS AND METHODS
In this study, Solid Dispense Unit (SDU) and Gravimetric Dispensing Unit for Powder fine
dosing (GDU‐Pfd), both of which are commercially available automated powder dispensing
tools from Chemspeed Technologies, were examined.10 These tools have been used for solid
dispensing in pharmaceutical laboratories. Both tools utilized gravity to deliver ChemBeads
from the dispensing bottle to the receiving vial. The SDU used rotary action via an auger to
dispense ChemBeads, whereas the GDU‐Pfd controls the dispense of ChemBeads via rapid
open/close actions of an opening at the bottom of the solid dispensing bottle.
ACS Paragon Plus Environment

Page 5 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
To ensure broad adoption of the ChemBead technology, manual distribution of
ChemBeads using calibrated scoops was also examined. A suite of 3D printed scoops with
calibrated scoop volume (calibrated to 5, 10 and 50 mg ChemBeads) were used to
distribute various solid reagents by simply scooping out Chembeads followed by lightly
tapping the scoop to remove any materials that were accumulated over the rim of the
scoop.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The inflated mass also enabled the delivery of sub‐milligram quantity of native reagents
with milligram amount of ChemBeads (at 1% loading, 5 mg ChemBeads contain 0.05 mg of
native solid reagent). This greatly simplified the solid dispensing process for both
automated and manual methods, i.e., handling milligrams quantity of ChemBeads vs
micrograms of native solid reagents.
Automated Solid Dispensing Platform and Calibrated Scoop Experiments.
The mass dispensed (mg) and dispense time (second) data were extracted from the log
files of the automated solid dispensing systems. Dispense time refers to the time required
for a given type of ChemBeads to dispense completely to a vial at the targeted mass, but
excluding the time required to move the dispense head into position prior to the actual
dispense. For the manual solid dispensing experiments using calibrated scoops, an
automated weighing Tecan EVO robot equipped with a calibrated balance was used to
determine the tare masses of a set of empty 4mL vials. ChemBeads were then dispensed
into the 4 mL vial using the calibrated scoops. The set of 4 mL vials were re‐weighed on the
Tecan EVO to determine the net mass of ChemBeads in each vial. The average dispense time
(second) per manual dispense using the calibrated scoops was defined as the difference
between the start and end time needed to dispense all 3 target masses in the vials, divided
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 22
1
2
3
4
5
6
7
8
9
by the total number of vials. The mean, median, standard deviation, % RSD calculations on
mass dispensed, and dispense time data were determined. The % error of each mass
dispensed was defined as the difference between the mass dispensed and the target mass,
divided by the target mass, and multiplying the result by 100 as shown below:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢁ
ꢃ
mass dispensed ꢂ target mass
% error ꢀ
ChemBeads Preparation.
x 100
target mass
Seven test solids were coated on commercially available glass or polystyrene beads. These
host materials are easily accessible, highly flowable, have automation friendly particle sizes
(200‐500 micron) and have a round smooth surface for uniform coating. All seven solids
and glass beads were obtained from Millipore Sigma and the polystyrene beads were
purchased from Advanced ChemTech (Table 1).
Table 1. The seven solids and the host materials investigated in this study.
Material
Grade
Lot number
SLBK4566V
SLBW6877
BCBP4505V
BCBM9216V
WXBC6499V
SLBV9983
D‐mannitol
ACS
fumed silica
Aggregate powder 0.2 – 0.3 µm
BioUltra
L‐proline
calcium carbonate
polyvinylpolypyrrolidone (PVPP)
sodium chloride
thiamine HCl
BioUltra
Average mol wt 40,000
BioXtra
BioReagent
PN: G1277,
212‐300 µm
PN: SP5070
35‐45 mesh
SLBW7025
glass beads
SLBS1081V
33344
polystyrene beads
D‐mannitol, L‐proline and sodium chloride appeared crystalline and chunky. They were
first milled down to a fine powder prior to the coating process on both glass and
polystyrene beads.7 The dry powder coating process was carried out using a Resonance
ACS Paragon Plus Environment

Page 7 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Acoustic Mixer (RAM).11 The RAM is a mixing instrument that relies upon the application of
a low‐frequency, high‐intensity acoustic field to facilitate mixing. A certain loading of the
solid reagent was added to a vial containing either glass or polystyrene beads and then
mixed using the RAM at 70% intensity for 5 to 15 minutes. For each solid material, we
investigated different percent loadings of the solid to the two host materials. ChemBeads
with maximum coating performance were used in this study (Table 2A and 2B). The
maximum loading was defined as the maximum amount of solid reagent that can be coated
on the glass and polystyrene beads evenly with no excess and loose chemical solids present.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2A. Maximum loadings of solids coated on glass beads.
ChemBead Mass Dispensed/(mg)
Loading on
glass beads
Solids
50
10
2
1
0.5
Estimated* Mass of Native Solid Dispensed (mg)
D‐mannitol
fumed silica
L‐proline
4.70%
1.70%
1.80%
4.90%
3.40%
2.35
0.85
0.90
2.45
1.70
0.47
0.17
0.18
0.49
0.34
0.094
0.034
0.036
0.098
0.068
0.047
0.017
0.018
0.049
0.034
0.0235
0.0085
0.009
sodium chloride
thiamine HCl
0.0245
0.017
*Amount calculated based on the % loading of the solids
Table 2B. Maximum loadings of solids coated on polystyrene beads.
ChemBead Mass Dispensed/(mg)
Loading on
Solids
polystyrene
beads
50
10
2
1
0.5
Estimated* Mass of Native Solid Dispensed (mg)
D‐mannitol
L‐proline
1.80%
4.00%
2.50%
20.0%
7.00%
1.80%
0.90
2.00
1.25
10.0
3.50
0.90
0.18
0.40
0.25
2.00
0.70
0.18
0.036
0.08
0.05
0.40
0.14
0.036
0.018
0.04
0.009
0.02
calcium carbonate
PVPP
0.025
0.20
0.0125
0.10
sodium chloride
thiamine HCl
0.07
0.35
0.018
0.009
*Amount calculated based on the % loading of the solids
Study Scope, Design, and Workflow
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 22
1
2
3
4
5
6
7
8
9
The study included the dispensing of ChemBeads to one vial as well as multiple vials
using the automated platforms and the calibrated scoops and, was designed to test practical
situations encountered mostly in HTE workflows by scientists in pharmaceutical
laboratories. The experiment was standardized and included the following aspects:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Target mass: 0.5, 1.0, 2.0, 10.0, and 50.0 mg were used for the automated solid dispensing
platforms. The 5, 10 and 50 mg calibrated scoops were used for manual dispensing of solid
reagents coated on glass beads; 5 mg scoop was the smallest size designed using a 3D
printer in our lab.12 The calibrated scoops for polystyrene beads are currently in
development.
Number of runs: 5 runs for each type of ChemBeads (5 coated on glass and 6 coated on
polystyrene beads) at a given target mass into a 20 mL vial for the automated solid
dispensing platform. 5 runs (5, 10, and 50 mg target mass) using the calibrated scoops into
20 separate 4 mL vials, for a total of 25 runs for all the ChemBeads using the automated
platforms and 15 runs for all the ChemBeads using the calibrated scoops.
Number of dispenses: 1 run = 20 dispenses of the ChemBeads, for a total of 500 dispenses
for each type of ChemBeads on the automated solid dispensing platforms and 300
dispenses using the calibrated scoops. A total of 7000 dispenses were made in the study.
Run order: The run involved 5 types of ChemBeads (glass beads) and 6 types of
ChemBeads (polystyrene beads) aligned consecutively with varied target masses (50.0,
10.0, 2.0, 1.0, and 0.5 mg) on the automated solid dispensing platforms. For the calibrated
scoops, the run involved scooping one type of ChemBeads at the different target masses (5,
10, and 50 mg consecutively) in the 20 separate 4 mL vials for each target mass.
ACS Paragon Plus Environment

Page 9 of 22
Organic Process Research & Development
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
9
The dry powder coating process of the seven solids on the beads resulted in maximum
loading of D‐mannitol, L‐proline, sodium chloride and thiamine HCl on both glass and
polystyrene beads, fumed silica on glass beads and calcium carbonate and PVPP on
polystyrene beads. The low effective mass of solid reagents (1‐20 %) on beads was
maximized for each individual solid (Table 2). For instance, dispensing 10 mg of thiamine
HCl ChemBeads at 3.4% loading will deliver 0.34 mg of native thiamine HCl. Purities and
quantities assessment of D‐mannitol, thiamine HCl and L‐proline ChemBeads were
confirmed using quantitative NMR. In addition, a HPLC based quantification method was
used to validate the amount of thiamine HCl coated on glass beads. In all cases examined,
loading accuracy was determined to be within ±10% of the calculated values. (see
Supporting Information for both quantitative NMR and HPLC based quantification results).
For NaCl, PVPP, CaCO3 and fumed silica, where NMR or HPLC quantification methods are
not applicable, we used mass difference of ChemBeads (before and after coating) to
estimate the loading of solids on beads.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Solid properties of each ChemBead were characterized by a variety of methods commonly
used in the pharmaceutical industry to evaluate their flow properties, density and
hygroscopicity (see Supporting Information for details). Results indicated that ChemBeads
have low variability of density, regardless of the solid reagent coated on the beads and
retained favorable flowability of the host material.
We verified the weighing data collected for any outliers and recurring patterns from the
7000 dispenses. A total of 90 dispenses were discounted and 6910 dispenses were used for
the analysis. The discounted dispenses were mostly due to user error in defining the
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 22
1
2
3
4
5
6
7
8
9
parameters on the GDU‐Pfd control software. There was no machine stall, machine time
out, or zero value observed in this study. The calibrated scoops were designed for coated
reagents on glass beads only and therefore they were only used to dispense D‐mannitol,
fumed silica, L‐proline, sodium chloride and thiamine HCl ChemBeads.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Scatterplot of % error values (‐80% to 80%) vs. dispense number (1 – 20),
broken down by reagents on glass beads (columns), target mass in mg (rows), and
instrument (colors).
ACS Paragon Plus Environment

Page 11 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Scatterplot of % error values (‐100% to 200%) vs. dispense number (1 – 20),
broken down by reagents on polystyrene beads (columns), target mass in mg (rows), and
instrument (colors).
Automated Solid Dispensing (SDU and GDU‐Pfd platforms)
All eleven ChemBeads were dispensed on both SDU and GDU‐Pfd equipped platforms. The
data retained for analysis is shown as a scatterplot of % error versus dispense number
through a run, broken down by the target mass, the type of techniques and the coated
material on glass beads (Figure 1) and on polystyrene beads (Figure 2).13 Solid dispensing
protocols for coated glass and polystyrene beads were developed separately on both SDU
and GDU‐Pfd equipped platforms. It is important to note that once the dispense protocols
were defined, the same solid dispensing protocol was used throughout the entire study,
regardless of the test subject, and without any additional optimization.
The % error values decreased for all seven ChemBeads as the targeted mass increased
from 0.5 mg to 50.0 mg (20 ‐ 80% for 0.5 mg to less than 5% for 50.0 mg). For the 50.0 mg
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 22
1
2
3
4
5
6
7
8
9
target mass, which translated to 0.5 mg solid reagent at 1% loading, and for 10.0 mg target
mass which translated to 0.1 mg solid reagent at 1% loading, the % errors were within 5%
across all eleven ChemBeads. A comprehensive statistical analysis (see Supporting
Information) of the 6910 dispenses also confirmed the accuracy at the 50 mg and 10 mg
level with a % RSD of less than 2% and 6% respectively. The dispensing at the 0.5 mg level
was performed to test the limitation of the technique. From Figures 1 and 2, some variation
in the behavior of the ChemBeads based on the solid loaded and a wider % error was
observed at the 0.5 mg level. The glass beads (diameter 212 – 300 µm) and polystyrene
beads (diameter 350 – 500 µm) used in this study were not uniform in diameters.
Dispensing at 0.5 mg and 1.0 mg represented a small number of beads (~40 glass
beads/mg and ~80 polystryene beads/mg) and therefore the variation in sizes became
significant and thus explaining the variation in behavior. When a higher mass of beads was
dispensed (≥ 2 mg), the number of beads became high enough that the variation in
diameter of individual bead played a less significant role in the dispense accuracy. We have
included the 0.5 mg and 1.0 mg dispenses data to test the limitation of this technique.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HTE WG found that native fumed silica has a very low bulk density, which made it difficult
to dispense using SDU.6 We demonstrated that fumed silica at 1.7% loading on glass beads
can be easily dispensed by both SDU and GDU‐Pfd equipped platforms at various target
masses with adequate accuracy (8 % error for 10.0 mg ChemBeads or 0.17 mg native solid,
2 % error for 50.0 mg ChemBeads or 0.85 mg native solid). When D‐mannitol, L‐proline,
sodium chloride and thiamine HCl were coated on both glass and polystyrene beads, similar
accuracy for all five target masses were observed. Overall, the GDU‐Pfd achieved slightly
better accuracy at all the target masses tested compared to the SDU. We speculated that the
ACS Paragon Plus Environment

Page 13 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
improved accuracy was the result of the newly designed solid dispensing bottle and the
GDU‐Pfd tool, which allow more user‐definable parameters (acceleration, amplitude, chunk
size, etc) and therefore provided better control of the solid dispensing mechanism.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Manual Solid Dispensing (Calibrated Scoops)
Data collected from the 1500 dispenses across the five types of ChemBeads using
calibrated scoops were analyzed and a scatterplot of % error versus dispense number
through a run, analyzed by the coated material dispensed and the target mass is shown in
Figure 1. A similar trend to the automated platforms was observed with % error values
decreased as the target mass values increased from 5 mg to 50 mg. When the 50 and 10 mg
calibrated scoops were used, less than 10% error was observed. The smallest scoop
available was the 5 mg calibrated scoop and rendered about 0.25 mg (at 5% loading on
beads) of native solid reagent at 20% error or less. Therefore, the 50 mg and 10 mg
calibrated scoops could be a general and efficient method for manual dispense of
ChemBeads without the need of an analytical balance or automated solid dispensing
platforms.
Comparative Study of the 3 Solid Dispensing Techniques
A comparative analysis of the three solid dispensing methods was carried out by
calculating % error for all the mass dispenses as shown in Figure 1. The % error values
decreased significantly for all the ChemBeads as the target mass values increased from 0.5
mg to 50.0 mg. In all cases, 5‐10% error was observed for target masses of 50.0 mg and
10.0 mg regardless of dispense method, and 2.0 mg target mass when using the GDU‐Pfd
tool. For our AbbVie internal HTE efforts, we have concluded that this level of accuracy is
adequate and does not affect the outcome of our screening efforts.
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 22
1
2
3
4
5
6
7
8
9
All three solid dispensing platforms were under different laboratory environments. The
SDU was located inside a constant nitrogen purged box and the GDU‐Pfd was inside a closed
environment (without nitrogen purge). The calibrated scoops were used in an open bench
environment. Overall, the laboratory environment had little or no impact on the % error
values observed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Dispense Time Consideration
In a high throughput experimentation laboratory, where hundreds of solid reagents will
need to be dispensed, efficiency of a platform in delivering targeted mass becomes a
practical consideration. We collected dispense times from the log files of each platform and
recorded the average dispense times for using the calibrated scoops. Dispense time and
mass dispensed data from each automated solid dispensing platform and calibrated scoops
were grouped by type of ChemBeads, target mass and different solid dispensing techniques,
and then quantified by calculating mean, median, standard deviation, and % RSD for each
subset (see Supporting Information).
ACS Paragon Plus Environment

Page 15 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Scatterplot of % error values (‐80% to 80%) vs. dispense time (0 – 70 s), broken
down by reagents on glass beads (columns), target mass in mg (rows), and instrument
(colors).
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 22
1
2
3
4
5
6
Figure 4. Scatterplot of % error values (‐100% to 200%) vs. dispense time (0 – 200 s),
broken down by reagents on polystyrene beads (columns), target mass in mg (rows), and
instrument (colors).
7
8
9
A scatterplot comparing the % error vs. dispense time values, analyzed by the target
masses, the type of dispensing techniques and the coated material on glass beads is shown
in Figure 3 and on polystyrene beads in Figure 4. Dispense time increased as the target
mass increased from 0.5 mg to 50.0 mg. A consistent trend in dispense time difference was
observed when comparing the two automated platforms for all ChemBeads (both glass and
polystyrene beads) across the different target masses with longer dispense time observed
on the GDU‐Pfd. For example, it took an average of 5 to 10 seconds on the SDU and 10 to 15
seconds on the GDU‐Pfd platform for each dispensing at the 0.5 mg level for solids coated
on glass beads. At the 50 mg target mass, it took an average of 10 to 15 seconds on the SDU
and 35 to 45 seconds on the GDU‐Pfd for each dispensing. The dispense time was also
longer (≥ 10 seconds) when dispensing polystyrene beads (D‐mannitol, L‐proline, sodium
chloride and thiamine HCl in Figure 4) mostly at the 50.0 and 10.0 mg target mass using the
GDU‐Pfd when compared to the same solids on glass beads. From the characterization data,
it was observed that polystyrene beads have a density approximately half that of glass
beads, potentially explaining their longer dispense time as larger volume of ChemBeads
will need to be dispensed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Using the calibrated scoops for manual dispensing demonstrated the shortest dispense
time among the three methods studied and comparable % error (±10%) for the 10 mg and
50 mg target masses (Figure 3). These scoops can be used for fast manual dispensing of
ChemBeads to deliver native solid at sub‐milligram scale. This confirmed that the calibrated
scoops were as effective as the automated platforms for solid dispensing and would be
ACS Paragon Plus Environment

Page 17 of 22
Organic Process Research & Development
1
2
3
4
5
6
practical for setting up high‐throughput experimentation when automation equipment was
not readily available.
7
8
9
Application of ChemBeads for Chemical Reactions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To demonstrate the utility of ChemBeads in setting up chemical reactions, two chemical
transformations using L‐proline and Boc‐L‐proline ChemBeads and their corresponding
native solid reagents were set up and the isolated yields recorded.
Scheme 1: Fmoc protection of L‐proline 114
Scheme 2: Acylation of 315
The comparative study of the reactivity between native L‐proline, Boc‐L‐proline and their
corresponding ChemBeads versions (both on glass and polystyrene beads) showed no
major difference in the % isolated yields (Table 3). These results indicated that the inert
nature of the glass and polystyrene beads do not interfere with the chemical reactivity of
the native solid reagents.
Table 3. Comparative study of chemical reactivity of the native solids and ChemBeads.
Chemical Transformation
Material
% Yield
native
89
Protection with L‐proline
1.8 % of L‐proline on glass ChemBeads
4.0 % of L‐proline on polystyrene ChemBeads
native
89
85
70
1.9 % of Boc‐L‐proline on glass ChemBeads
67
Acylation with Boc‐L‐proline
4.0 % of Boc‐L‐proline on polystyrene
67
ChemBeads
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 22
1
2
3
4
CONCLUSIONS AND OUTLOOK
5
6
7
8
9
SDU, GDU‐Pfd, as well as the calibrated scoops were successfully applied to dispense
different types of ChemBeads with comparable mean mass values and relatively consistent
dispense time for each target mass examined. This study proved that the ChemBeads
technology is a practical approach to dispense sub‐milligram amounts of solid reagents
using either automation platforms or calibrated scoops. The level of accuracy for the
different target masses was relatively consistent for the seven types of ChemBeads at 2.0,
10.0 and 50.0 mg using the automated solid dispensing platforms. The calibrated scoops
demonstrated similar and consistent accuracy for the 10.0 and 50.0 mg target mass with
shorter dispense time, validating its general utility as a fast and effective manual solid
dispensing method on the bench without the need of an analytical balance and automation
equipment.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
With this novel technology, over 450 ChemBeads of different solid reagents have been
prepared and used exclusively in our internal HTE efforts with great success. As shown in
this study, the process was straightforward, general and applicable to different solid
reagents, regardless of its physical and chemical nature. The uniform physical property of
the ChemBeads, regardless of the original form of the native solid reagent, facilitated solid
dispensing at low targeted masses using the GDU‐Pfd tool. With the ability to deliver sub‐
milligram quantities of the native solid reagents in vials, screening multiple reaction
conditions in parallel can be done without exhausting precious starting materials. No major
difference in target mass accuracy was observed for dispensing solid reagents coated on
glass or polystyrene beads using both SDU and GDU‐Pfd platforms. After the 6910
dispenses recorded, the material was observed to remain on the beads throughout the
ACS Paragon Plus Environment

Page 19 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
dispensing process. We also demonstrated that we could use the calibrated scoops with the
ChemBead technology to manually dispense sub‐milligram quantities of solid reagents
depending on their loadings using the 5, 10 and 50 mg scoops. These calibrated scoops
with the ChemBeads technology would be an asset to increase applications of HTE in
laboratories not equipped with automated solid dispensing platforms. The inert nature of
the glass and polystyrene beads did not interfere with the reactivity of the solid reagents
coated on them hence rendering them suitable to set up microscale high‐throughput
experimentation. When choosing a host material (glass or polystyrene) for coating, glass
beads should be the first choice due to their availability from multiple vendors, lower price
and solvent compatibility. Polystyrene beads should be considered as host when glass
beads do not achieve the desired reagent loading.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ChemBeads technology, when coupled with a suitable solid dispensing platform, would be
a solution for the development of a single solid dispensing protocol for solid reagents of
various physical properties. We fully anticipate that the ChemBead technology will find
broad application in both academic and industrial laboratories
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at
DOI:
Chemspeed Swave and Chemspeed Swing Isynth solid dispensing platforms; Calibrated
scoops; Dry powder coating using Resonant Acoustic Mixer; Tables containing
dispensed data using Chemspeed Swave and Swing Isynth and calibrated scoops;
quantitative NMR data for thiamine HCl and D‐mannitol; Quantification data for
thiamine HCl using calibration curve; Tables of mean mass and % RSD mass values from
dispenses on the Chemspeed Swave and Swing Isynth as well as calibrated scoops;
Physical characterization data of ChemBeads.
AUTHOR INFORMATION
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 22
1
2
3
4
Corresponding Author: Noah P. Tu
Email: noah.tu@abbvie.com. Tel.: +1 (847) 938 1774
ORCID
Noah P. Tu: 0000‐0002‐9164‐3815
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Address: 1 N Waukegan Road, North Chicago, IL 60064, United States
Disclosures: All authors are employees of AbbVie. The design, study conduct, and financial
support for this research were provided by AbbVie. AbbVie participated in the
interpretation of data, review, and approval of the publication.
Funding: This research was sponsored by AbbVie.
Competing interests: None declared.
Data and materials availability: All data is available in the manuscript or the
supplementary materials.
ACKNOWLEDGMENTS
We thank Philip Searle for data analysis and visualizations, Rick Yarbrough for
quantitative NMR studies and Raimundo Ho for helpful discussions on ChemBeads
characterization experiment.
REFERENCES
(1) Hughes, J. P.; Rees, S.; Kalindjian, S. B.; Philpott, K. L. Principles of early drug discovery. British Journal
of Pharmacology. 2011, 162, 1239‐1249.
(2) Collins, K. D.; Gensch, T.; Glorius, F. Contemporary screening approaches to reaction discovery and
development. Nat. Chem. 2014, 6, 859 ‐871.
(3) (a) Perera, D.; Tucker, J. W.; Brahmbhatt, S.; Helal, C. J.; Chong, A.; Farrell, W.; Richardson, R.; Sach, N.
W. A platform for automated nanomole‐scale reaction screening and micromole‐scale synthesis in
flow. Science. 2018, 359, 429‐434. (b) Krska, S. W.; DiRocco, D. A.; Dreher, S. D.; Shevlin, M. The
evolution of chemical high‐throughput experimentation to address challenging problems in
pharmaceutical synthesis. Acc. Chem. Res. 2017. 50, 2976‐2985. (c) Buitrago Santanilla, A.; Regalado,
E. L.; Pereira, T.; Shevlin, M.; Bateman, K.; Campeau, L‐C.; Schneeweis, J.; Berritt, S.; Shi, Z‐C.;
Nantermet, P.; Liu, Y.; Helmy, R.; Welch, C. J.; Vachal, P.; Davies, I. W.; Cernak, T.; Dreher, S. D.
Nanomole‐scale high‐throughput chemistry for the synthesis of complex molecules. Science. 2015,
347, 49‐53.
(4) (a) University of Pennsylvania Department of Chemistry High Throughput Screening Center.
https://www.chem.upenn.edu/content/penn‐chemistry‐upennmerck‐high‐throughput‐
experimentation‐laboratory (accessed Feb 08, 2019). (b) Caltech Division of Chemistry and Chemical
Engineering 3CS. http://chche.divisions.caltech.edu/facilities/research‐centers/center‐for‐catalysis‐
and‐chemical‐synthesis‐3cs (accessed Feb 08, 2019). (c) Princeton University Department of
Chemistry Merck Catalysis Center. https://chemistry.princeton.edu/research‐facilities/merck‐
catalysis‐center (accessed on Feb 08, 2019).
(5) (a) Shevlin, M. Practical high‐throughput experimentation for chemists. ACS Med. Chem. Lett. 2017, 8,
601‐607. (b) Selekman, J. A.; Qiu, J.; Tran, K.; Stevens, J.; Rosso, V.; Simmons, E.; Xiao, Y.; Janey, J. High‐
throughput automation in chemical process development. Annu Rev Chem Biomol Eng. 2017, 8, 525‐
547. (c) Cernak et al. Microscale high‐throughput experimentation as an enabling technology in drug
discovery: application in the discovery of (piperidinyl)pyridinyl‐1H‐benzimidazole diacylglycerol
acyltransferase 1 inhibitors. J. Med. Chem. 2017, 60, 3594‐3605.
ACS Paragon Plus Environment

Page 21 of 22
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
(6) Bahr, M. N.; Damon, D. B.; Yates, S. D.; Chin, A. S.; Christopher, J. D.; Cromer, S.; Perrotto, N.; Quiroz, J.;
Rosso, V. Collaborative Evaluation of Commercially Available Automated Powder Dispensing
Platforms for High‐Throughput Experimentation in Pharmaceutical Applications. Org. Process Res. Dev.
2018, 22, 1500‐1508.
(7) Tu, N. P.; Dombrowski, A. W.; Goshu, G. M.; Vasudevan, A.; Djuric, S. W.; Wang, Y. High throughput
reaction screening with nanomoles of solid reagents coated on glass beads. Angew. Chem. Int. Ed.
2019, 58, 7987‐7991.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(8) Dombrowski, A.; Tu, N. P.; Wang, Y. High throughput methods for screening chemical reactions using
reagent‐coated bulking agents. US20190033185, 2019.
(9) (a) Dahmash, E. Z.; Mohammed, A. R. Functionalized particles using dry powder coating in
pharmaceutical drug delivery: promises and challenges. Expert Opin Drug Deliv. 2015, 12, 1867‐1879.
(b) Mullarney, M. P.; Beach, L. E.; Davé, R. N.; Langdon, B. A.; Polizzi, M.; Blackwood, D. O. Applying dry
powder coatings to pharmaceutical powders using a comil for improving powder flow and bulk
density. Powder Technology. 2011, 212, 397‐402. (c) Tomas, J. Kleinschmidt, S. Improvement of
flowability of fine cohesive powders by flow additives. Chem. Eng. Technol. 2009, 32, 1470‐1483. (d)
Yang, J.; Sliva, A.; Banerjee, A.; Dave, R. N.; Pfeffer, R. Dry particle coating for improving the flowability
of cohesive powders. Powder Technology, 2005, 158, 21‐33. (e) Pfeffer, R.; Dave, R. N.; Wei, D.;
Ramlakhan, M. Synthesis of engineered particulates with tailored properties using dry particle
coating. Powder Technology. 2001, 117, 40‐67. (f) Ramlakhan, M.; Yu Wu, C.; Watano, S.; Dave, R. N.;
Pfeffer, R. Dry particle coating using magnetically assisted impaction coating: modification of surface
properties and optimization of system and operating parameters. Powder Technology. 2000, 112, 137‐
148.
(10) Chemspeed Technologies Robotic Tools. https://www.chemspeed.com/technologies/robotic‐tools‐
features/ (accessed on Apr 22, 2019).
(11) ResodynTM Acoustic Mixers. https://resodynmixers.com/ (accessed on Feb 08, 2019).
(12) 3D Systems. https://www.3dsystems.com/shop/support/cubepro/videos (accessed on Apr 22,
2019).
(13) For ease of comparison, data representation formats were kept as closely as possible to HTE WG.
(14) Popovici‐Muller et al. Discovery of AG‐120 (Ivosidenib): a first‐in‐class mutant IDH1inhibitor for the
treatment of IDH1 mutant cancers. ACS Med. Chem. Lett., 2018, 9, 300‐305.
(15) General Procedure used for acylation reaction: To three 4 mL vials each containing Boc‐L‐proline 3
(native solid, 1.9 % coated on glass beads and 4.5 % coated on polystyrene beads, 0.12 mmol, 1.0
equiv.) was added HATU (0.17 mmol, 1.5 equiv.), DIEA (0.23 mmol, 2.0 equiv.) and acetonitrile (0.1 M)
and allowed to shake using an orbital heater shaker for 5 min at 60°C, followed by addition of p‐
tolylmethanamine 4 (0.17 mmol, 1.5 equiv.) and allowed to shake overnight. Upon completion water
was added to the resulting mixture and extracted with ethyl acetate. The combined extracts were
dried down using Biotage® V‐10 and treated with TFA in methanol to remove the Boc group. The
crude product was dissolved in 1.5 mL DMSO/H2O (1:1), filtered, and the solid washed with additional
1.5 mL of DMSO. The crude product was purified using reverse phase preparative HPLC to yield pure
1
(S)‐N‐(4‐methylbenzyl)pyrrolidine‐2‐carboxamide 5 as a colorless oil. H NMR (400 MHz, CDCl3) δ:
8.36 (t, J = 5.7 Hz, 1H), 4.58 (d, J = 9.5 Hz, 1H), 4.33 (d, J = 5.8 Hz, 2H), 3.38 – 3.22 (m, 2H), 2.30 (s, 4H),
1.95 (pt, J = 9.0, 4.8 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ: 168.6, 137.2, 134.4, 129.3, 127.4, 59.6, 46.4,
43.6, 30.1, 24.4, 21.0.
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment