Portable open-source autosampler for shallow waters

Abstract

Automated water sampling can be very useful, but open-source choices are limited. Here I present an autosampler which consists of a gantry robot that delivers water from a syringe pump to 24 capped 40 ml vials. The autosampler is controlled using an Android tablet automatized using Macrodroid. Three rinsing cycles ensure negligible carryover between consecutive samples. Hourly sampling from a creek under rainy conditions suggested that total organic carbon in water was diluted by the rain. Some important limitations: 1) the autosampler must be on a steady, flat, horizontal surface; 2) unattended sampling can only last as long as the batteries powering the tablet and the motors; 3) distance from the syringe pump to water cannot exceed ~2 m in height and ~4 m in length for 3 mm tubing; 4) sampling frequency does not exceed one sample every eleven minutes. However, because of its open design, the autosampler can be modified and improved to not only overcome these limitations, but also potentially expand its scope to more demanding sampling if necessary.

Full text

Portable open-source autosampler for shallow waters
Matheus C. Carvalho
Centre for Coastal Biogeochemistry Research, School of Environment, Science and Engineering. Southern Cross University, PO box 157, Lismore 2480, NSW, Australia
article info
Article history:
Received 28 April 2020
Received in revised form 26 August 2020
Accepted 4 September 2020
Keywords:
3D printing
Autosampler
Android, Arduino
Field sampling
Liquid handling
OpenSCAD
Macrodroid, Marlin
Stable isotopes
Syringe
Water sampling
abstract
Automated water sampling can be very useful, but open-source choices are limited. Here I
present an autosampler which consists of a gantry robot that delivers water from a syringe
pump to 24 capped 40 ml vials. The autosampler is controlled using an Android tablet
automatized using Macrodroid. Three rinsing cycles ensure negligible carryover between
consecutive samples. Hourly sampling from a creek under rainy conditions suggested that
total organic carbon in water was diluted by the rain. Some important limitations: 1) the
autosampler must be on a steady, flat, horizontal surface; 2) unattended sampling can only
last as long as the batteries powering the tablet and the motors; 3) distance from the
syringe pump to water cannot exceed ~2 m in height and ~4 m in length for 3 mm tubing;
4) sampling frequency does not exceed one sample every eleven minutes. However,
because of its open design, the autosampler can be modified and improved to not only
overcome these limitations, but also potentially expand its scope to more demanding
sampling if necessary.
Ó2020 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Specifications table
Hardware name Iara
Subject area Environmental, Planetary and Agricultural Sciences
Hardware type Portable autosampler for discrete water samples
Open Source License GNU General Public License (GPL) 3.0
Cost of Hardware AU$850
Source File Repository https://osf.io/2c6t5/ or https://doi.org/10.17605/OSF.IO/2C6T5
1. Hardware in context
Understanding the natural environment is a combination of measurements and modelling. Although modelling provides
invaluable insights for this purpose [1], it will only be useful as long as good data underlying the rationales are available [2].
Data are gathered through samples, and thus samples are at the core of good environmental research.
Collecting data can be time-consuming, tiresome, tedious, and even dangerous, depending on the kind of sample. There-
fore, it is no wonder that automated data collection has been made available through sensors attached to loggers [3–11].
https://doi.org/10.1016/j.ohx.2020.e00142
2468-0672/Ó2020 Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
E-mail address: mcarvalh@scu.edu.au
HardwareX 8 (2020) e00142
Contents lists available at ScienceDirect
HardwareX
journal homepage: www.elsevier.com/locate/ohx

However, there are still many kinds of measurements that cannot be performed in situ, and that demand that the samples are
collected and brought to a laboratory for further analysis. This is very common for water samples, for example.
Automated water sampling for further analysis is commercially available [12–14] but, as most commercial scientific
equipment, these autosamplers are rather costly and difficult to customize due to their closed, patented designs. As an alter-
native, an open-source autosampler has been presented [15]. It allows the collection of two water aliquots, and can be
deployed under water. If more than two water aliquots are necessary, more units of the autosampler can be deployed. Here
I present a different kind of autosampler that allows the collection of a larger number (in the example provided here, 24) of
samples. It stays outside of the water, and can be particularly useful for shallow waters. It is inspired on open-source
autosamplers for laboratory use [16,17], with modifications to allow portability.
2. Hardware description
The autosampler consists of a gantry robot using a thick (18 gauge) syringe needle as an end effector (Fig. 1). This needle
is moved on straight lines on three axes: X (horizontal), Y (horizontal) and Z (vertical).
The needle is connected to a large (50 ml) syringe pump using 1/8
00
Teflon and 3 mm vinyl tubing, in a design similar to
previous ones [18–20]. The pump is filled and emptied by the movement of the syringe plunger on a fourth (vertical) axis, E.
In between the pump and the needle, there is a three-way valve, which diverts the flow allowing for sampling and rinsing.
The pump delivers the water to the needle, which can be moved to the sample vials placed inside a tray fixed to the autosam-
pler body, or the drain (a simple pipe, in this case). The whole setup can be placed inside a large storage box.
The gantry robot and the syringe pump are actuated by stepper motors, while the valve by a servo motor. A rechargeable
and portable 12 V battery powers the motors. The motors are controlled using an MKS Gen 1.4 control board, which has the
Marlin firmware installed [16,17,21,22]. The board is connected to a tablet running on the Android operating system, on
which Macrodroid [23–25], a scripting interface, is used to program the motor movements. This is a similar approach to
the use of AutoIt to integrate analytical instruments [26,27]. A communication software, Serial USB Terminal [28,29], is used
to send the commands, while allowing manual input if needed.
Videos of the autosampler working are shown in Supplementary Information 1, 2 and 3.
3. Design files
Although the autosampler is built using mostly off-the-shelf parts, others need to be 3D-printed. All parts were printed
using 1.75 mm PLA filament, with layer height of 0.3 mm, shell thickness of 0.8 mm, fill density of 90% (important to ensure
part’s strength), nozzle temperature of 195 °C, bed temperature of 50 °C, support everywhere and raft as a platform for adhe-
sion. All files can also be accessed on https://doi.org/10.17605/OSF.IO/2C6T5.
3.1. Design files summary
Fig. 1. Portable autosampler for discrete surface water samples.
M.C. Carvalho HardwareX 8 (2020) e00142
2

Design file name File type Open source license Location of the file
Stepper-motor plate (horizontal X axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Stabilizer plate (X axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Plate spacer (X axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Wheel spacer (horizontal X, Y axes) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Stepper-motor plate (Y axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Stepper-motor mount (vertical Z and E axes) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Linear carrier (Z axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Syringe holder (E axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Plunger holder (E axis) OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Servo motor mount OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Servo-valve connector 1 OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Servo-valve connector 2 OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Sample tray OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Sample tray cover OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Sample tray spacer OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
Spacer for control board OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
45°Pipe holder OpenSCAD GNU General Public License (GPL) 3.0 https://osf.io/
2c6t5/
4. Bill of materials
Notes:
1) Many items are listed in excess
2) The aluminum extrusion profiles can be bought in different sizes and with the M6 thread already made, depending on
the supplier. Alternatively, they can be cut and threaded after purchase.
3) Prices for 3D-printed items were calculated using a price of AU$20 for a 1 kg roll, and the estimate of material used by
each item done by the 3D-printing software.
Figure or
reference
Component Quantity Cost
per unit
–AU$
Total
cost –
AU$
Source of materials Material
type
Fig. 2 T slot 20 20 400 mm 4 $9.00 $36.00 https://www.aliexpress.com/
item/32979181065.html
Aluminum
Fig. 2 T slot 20 20 250 mm 2 $7.00 $14.00 https://www.aliexpress.com/
item/32979181065.html
Aluminum
(continued on next page)
M.C. Carvalho HardwareX 8 (2020) e00142
3

(continued)
Figure or
reference
Component Quantity Cost
per unit
–AU$
Total
cost –
AU$
Source of materials Material
type
Figs. 4 and 5 T slot 20 20 150 mm 1 $5.00 $5.00 https://www.aliexpress.com/
item/32979181065.html
Aluminum
Fig. 5 T slot 20 20 120 mm 3 $5.00 $15.00 https://www.aliexpress.com/
item/32979181065.html
Aluminum
Figs. 4 and 5 T slot 20 20 200 mm 1 $6.00 $6.00 https://www.aliexpress.com/
item/32979181065.html
Aluminum
Fig. 18 T slot 20 40 250 mm,
M6 threaded in holes
1 $7.50 $7.50 https://www.aliexpress.com/
item/32971223475.html
Aluminum
Fig. 4 Lead screw set, 150 mm 1 $3.78 $3.78 https://www.aliexpress.com/
item/32917521964.html
Stainless
steel,
brass
Figs. 14, 15 Lead screw set, 100 mm 1 $3.14 $3.14 https://www.aliexpress.com/
item/32917521964.html
Stainless
steel,
brass
Figs. 4, 5 NEMA 17 stepper motor
(set with 5)
1 $80.00 $80.00 https://www.aliexpress.com/
item/32376023464.html
Metal,
plastic
Fig. 31 Cables for NEMA 17
stepper motor (set with 4)
1 $6.00 $6.00 https://www.aliexpress.com/
item/32908906012.html
Metal,
plastic
Figs. 3, 15 Shaft coupler, 5 mm to
8 mm (set with 2)
1 $9.00 $9.00 https://www.aliexpress.com/
item/4000433138641.html
Metal
Figs. 2, 5 2028 bracket (set with 20) 1 $16.00 $16.00 https://www.aliexpress.com/
item/32642711369.html
Metal
Fig. 5 2525 bracket 1 $0.70 $0.70 https://www.bunnings.com.
au/carinya-25-x-25-x-20-x-
2mm-make-a-bracket_
p3960586
Metal
Fig. 12 Bearing for leadscrew (set
with 10)
1 $4.00 $4.00 https://www.aliexpress.com/
item/32925867143.html
Metal
Figs. 6, 7, 8 Nylon wheels (set with
10)
2 $6.50 $13.00 https://www.aliexpress.com/
i/32816586242.html
Metal,
nylon
Figs. 8, 18 GT2 timing belt, 6 mm
width, 2 m roll
1 $3.00 $3.00 https://www.aliexpress.com/
item/32820590582.html
Plastic
Figs. 8, 16 Pulley for GT2, 6 mm
width, 20 teeth, 5 mm
inner bore, set with 2
1 $7.00 $7.00 https://www.aliexpress.com/
item/32995102911.html
Metal
Figs. 2, 4 M5 screw, 10 mm length,
set with 20
3 $7.00 $21.00 https://www.aliexpress.com/
item/33007566910.html
Metal
Figs. 9, 20 M5 screw, 15 mm length,
set with 20
1 $7.50 $7.50 https://www.aliexpress.com/
item/33007566910.html
Metal
Figs. 6, 7 M5 screw, 30 mm length,
set with 20
1 $8.50 $8.50 https://www.aliexpress.com/
item/33007566910.html
Metal
Figs. 2, 6, 8, 29 M5 hammer nut, set with
50
2 $12.00 $24.00 https://www.aliexpress.com/
item/4000406680181.html
Metal
Optional M5 washer, set with 50 1 $2.50 $2.50 https://www.aliexpress.com/
item/32975752411.html
Metal
Fig. 20 M5 T nut, set with 30 1 $10.00 $10.00 https://www.ebay.com.au/
itm/10Pcs-T-Sliding-
Hammer-Nut-Block-Square-
Nuts-M4-M5-M6-Nut-20–30-
40–45-Ser-ex/324090034202
Metal
Fig. 19 M6 screw, 12 mm, set
with 20
1 $10.00 $10.00 https://www.aliexpress.com/
item/33006640055.html
Metal
Fig. 17 M6 screw, 70 mm, set
with 10
1 $10.00 $10.00 https://www.aliexpress.com/
item/33006640055.html
Metal
M.C. Carvalho HardwareX 8 (2020) e00142
4

(continued)
Figure or
reference
Component Quantity Cost
per unit
–AU$
Total
cost –
AU$
Source of materials Material
type
Fig. 21 M6 screw, 120 mm, set
with 10
1 $12.00 $12.00 https://www.aliexpress.com/
item/33006640055.html
Metal
Figs. 17, 21 M6 nuts and washers, set
with 10
2 $5.00 $10.00 https://www.aliexpress.com/
item/33006640055.html
Metal
Figs. 8, 15, 16 M3 screw, 10 mm, set
with 50
1 $3.50 $3.50 https://www.aliexpress.com/
item/4000193782706.html
Metal
Figs. 9, 10, 11,
13, 22, 28, 30
M3 screw, 25 mm, set
with 50
1 $5.00 $5.00 https://www.aliexpress.com/
item/4000193782706.html
Metal
Figs. 9, 10, 11,
13, 22, 28
M3 nuts, set with 25 2 $1.00 $2.00 https://www.aliexpress.com/
item/32977174437.html
Metal
Fig. 30 M3 hammer nuts, set with
20
1 $3.00 $3.00 https://www.aliexpress.com/
item/32847391903.html
Metal
Figs. 13, 15 18 G, 1.5 in. needle, set
with 5
1 $2.95 $2.95 https://www.ebay.com.au/
itm/TERUMO-Hypodermic-
Needles-18G-19G-21G-22G-
23G-25G-27G-30G-Premium-
Medical-Grade/
124113089538
Stainless
steel,
plastic
Fig. 21 Plastic clipboard, A4 size 1 $2.50 $2.50 https://www.
officeworks.com.au/shop/
officeworks/p/keji-clipboard-
a4-clear-plastic-kepla4cbcr
Plastic
Fig. 30 MKS Gen 1.4 control
board
1 $40.00 $40.00 https://www.aliexpress.com/
item/32935055346.html
Metal,
plastic
Fig. 30 A4988 stepper motor
driver
2 $3.00 $6.00 https://www.aliexpress.com/
item/4000590332485.html
Metal
Fig. 30 Drv8825 stepper motor
driver
2 $3.00 $6.00 https://www.aliexpress.com/
item/4000590332485.html
Metal
Figs. 27, 28 Servo motor 1 $11.00 $11.00 https://www.ebay.com.au/
itm/1–2-4PCS-MG996R-55g-
Metal-Gear-Torque-Digital-
Servo-for-RC-Helicopter-Car-
Robot/264267504263
Metal,
plastic
Figs. 27, 28 3-way valve, 1/8
00
1 $185.00 $185.00 https://www.swagelok.com/
en/catalog/Product/
Detail?part = SS-41GXS2
Metal,
plastic
Fig. 33 12 V, 12Ah battery 1 $49.00 $49.00 www.batteriesdirect.com.au/
shop/product/10174/rm12-
12.html
Metal
Figs. 1, 33 Android tablet 1 $97.00 $97.00 https://www.
officeworks.com.au/shop/
officeworks/p/lenovo-tab-e7-
7-16gb-tablet-za400039au-
syl7tabe7b
Metal,
plastic
Fig. 1 USB cable, 1 m 1 $5.00 $5.00 https://www.aliexpress.com/
item/32994837483.html
Metal,
plastic
Fig. 1 OTG USB cable 1 $1.50 $1.50 https://www.aliexpress.com/
item/4000202709849.html
Metal,
plastic
Fig. 1 Electric cable, 1 m 1 $2.00 $2.00 https://www.aliexpress.com/
item/32265973299.html
Metal,
plastic
Fig. 33 Storage box, 90 L capacity 1 $15.00 $15.00 https://
www.supercheapauto.com.
au/p/sca-sca-storage-box-90-
litre/580786.html
Plastic
(continued on next page)
M.C. Carvalho HardwareX 8 (2020) e00142
5

(continued)
Figure or
reference
Component Quantity Cost
per unit
–AU$
Total
cost –
AU$
Source of materials Material
type
Figs. 27, 28 1/8
00
PTFE tube, 1 m 1 $13.00 $13.00 https://www.ebay.com.au/
itm/1–8-OD-3–18mm-ID-1-
68mm-PTFE-Tubing-Tube-
Pipe-hose-per-meter-1m-
Length/323905143967
PTFE
Figs. 27, 28 1/8
00
fitting, male and
female set
2 $5.00 $10.00 https://www.aliexpress.com/
item/32838411990.html
Stainless
steel
Fig. 13 3 mm to male luer lock
fitting
1 $1.00 $1.00 https://www.aliexpress.com/
item/32994503237.html
Plastic
Fig. 32 3 mm tube, 5 m roll 1 $6.50 $6.50 https://www.bunnings.com.
au/pope-3mm-x-5m-clear-
vinyl-tubing_p3130556
Vinyl
Fig. 12 50 ml syringe 1 $10.00 $10.00 https://www.ebay.com.au/
itm/Syringes-TERUMO-1ml-
3ml-5ml-10ml-20ml-30ml-
50ml-Suit-Luer-Slip-Lock-
Syringe/143551845718
Plastic
Fig. 22 Pipe, 20 mm, 1 m 1 $4.00 $4.00 https://www.bunnings.com.
au/holman-20mm-x-1m-pvc-
pressure-pipe_p4750047
PVC
Items below are 3D-printed
Fig. 16 Stepper-motor plate (X
axis)
1 $2.00 $2.00 Section 3.1 PLA
Fig. 16 Stabilizer plate (X axis) 1 $2.00 $2.00 Section 3.1 PLA
Fig. 17 Plate spacer (X axis) 4 $0.10 $0.40 Section 3.1 PLA
Figs. 6, 7, 17 Wheel spacer (X, Y axes) 16 $0.02 $0.32 Section 3.1 PLA
Figs. 7, 8 Stepper-motor plate
(Y axis)
2 $2.00 $4.00 Section 3.1 PLA
Figs. 3, 4, 14, 15 Stepper-motor mount (Z,
E axis)
2 $0.35 $0.70 Section 3.1 PLA
Fig. 13 Linear carrier (Z axis) 1 $0.35 $0.35 Section 3.1 PLA
Fig. 12 Syringe holder (E axis) 1 $1.00 $1.00 Section 3.1 PLA
Fig. 9 Plunger holder (E axis) 1 $0.40 $0.40 Section 3.1 PLA
Figs. 28, 29 Servo motor mount 1 $0.40 $0.40 Section 3.1 PLA
Figs. 28, 29 Servo-valve connector 1 1 $0.10 $0.10 Section 3.1 PLA
Figs. 28, 29 Servo-valve connector 2 1 $0.10 $0.10 Section 3.1 PLA
Fig. 21 Sample tray 1 $2.00 $2.00 Section 3.1 PLA
Fig. 21 Sample tray cover 1 $1.00 $1.00 Section 3.1 PLA
Fig. 21 Sample tray spacer 8 $0.10 $0.80 Section 3.1 PLA
Fig. 30 Spacer for control board 2 $0.03 $0.06 Section 3.1 PLA
Fig. 22 45°Pipe holder 1 $0.20 $0.20 Section 3.1 PLA
5. Build instructions
5.1. Assembling the frame
Connect the 400 mm 20 mm 20 mm extrusion profiles to the 250 mm ones using the corner brackets (Fig. 2). Use
10 mm M5 screws and hammer nuts to fix the 2028 brackets to the profiles. It is necessary that the structure sits along a
plane, that is, it is flat. The brackets used here are useful for this purpose.
Connect the 150 mm lead screw to a stepper motor using the shaft coupler, and attach the motor to the motor mount
(Fig. 3; motor mount for two vertical axes, Z and E).
Connect the motor mount to a 200 mm 20 mm 20 mm extrusion profile (Fig. 4) using 10 mm M5 screws.
Connect the vertical profiles (120 and 200 mm ones, this last one already with the motor mount) using corner brackets
(Fig. 5).
M.C. Carvalho HardwareX 8 (2020) e00142
6

5.2. Assembling the Y axis
Arrange wheels, spacers, 30 mm M5 screws and nuts (Fig. 6; a hammer nut can substitute the one shown in the figure).
Attach the wheels and one motor to the Y axis motor mounts (Fig. 7). Attach one pulley to the motor shaft, aligning it to
the wheels (misalignment here leads to poor performance of the mechanism; see also Fig. 8 for an example of good
alignment).
Slide 400 mm extrusion profiles through the wheels on each plate and attach them to the frame (Fig. 8). On the one that
has the motor, wrap the timing belt around the pulley and beneath the wheels, then fix it to the profile using 10 mm M5
screws and hammer nuts. Notice the orientation of the plates, with the three upper holes further from the E axis motor.
Fig. 2. Horizontal base of the frame.
Fig. 3. E axis motor with its lead screw connected using a shaft coupler.
M.C. Carvalho HardwareX 8 (2020) e00142
7

5.3. Assembling the syringe driver (E axis)
Make holes in the syringe plunger to match those in the linear carriage and fix the syringe plunger to the linear carriage
(Fig. 9) using 15 mm M5 screws. Depending on the syringe model, the plunger can be fragile, so it may be better to drill the
Fig. 4. Motor mount attached to 200 mm extrusion profile.
Fig. 5. Assembled frame, including the E axis with its motor.
M.C. Carvalho HardwareX 8 (2020) e00142
8

holes initially with a rotary tool equipped with a thin drill bit, and then increase hole diameter with wider bits manually.
Attach a leadscrew nut to the plunger holder using M3 screws.
Slide the linear the plunger holder around the leadscrew (Fig. 10).
Make 3 mm holes on the syringe body matching those in the leadscrew support (Fig. 11). Fix the syringe to the support
(Fig. 11) using M3 screws.
Fix the E axis syringe holder to the extrusion profile (Fig. 12) using 10 mm M5 screws. Add a bearing around the leadscrew
in the appropriate slot in the support.
5.4. Assembling the Z axis
Attach a leadscrew nut to the Z axis linear carrier using M3 screws, and fix the luer – 3 mm connector also using a M3
screw. Attach the syringe needle to the luer connector (Fig. 13).
Attach the 100 mm leadscrew to the stepper motor shaft (Fig. 14). Attach a stepper motor to the Z axis motor mount and
the motor mount to the 150 mm profile (Fig. 14).
Slide the carrier around the leadscrew (Fig. 15).
5.5. Making the X axis
Attach wheels and motor to the X axis motor mount and support plate (Fig. 16). The procedure is the same as that for the
Y axis (Fig. 7).
Fig. 6. Wheel set to make the sliding motor carriers on axes X and Y.
Fig. 7. Motor mounts for Y axis.
M.C. Carvalho HardwareX 8 (2020) e00142
9

Connect the X axis plates using the 70 mm M6 screws and appropriate spacers (Fig. 17).
Slide the connected plates on the 20 40 mm extrusion profile and attach the timing belt in the same way as done for
the X axis (Fig. 18).
Connect the X axis to the Y axis plates using 12 mm M6 screws (Fig. 19). Use the two holes further from the syringe
(Fig. 19).
Connect the Z axis to the X axis plate using T nuts and 15 mm M5 screws (Fig. 20).
Fig. 8. Frame including the Y axis.
Fig. 9. Plunger attached to its holder.
M.C. Carvalho HardwareX 8 (2020) e00142
10

5.6. Making and placing the sample tray
Cut the clipboard to make it 190 mm long, so that it becomes a rectangle measuring 190 225 mm. Eight holes will be
needed, four with 5 mm in diameter, and the other 4 with 6 mm in diameter (Fig. 21, subfigure 1). The 6 mm holes should be
positioned 143 mm from each other on the width, and 185 mm on the length. The 5 mm holes should be positioned 175 mm
from each other on the width, and can be positioned 100 mm from each other on the length, although this distance is not
crucial.
Fig. 10. Plunger holder on the E axis.
Fig. 11. Syringe attached to E axis syringe holder.
M.C. Carvalho HardwareX 8 (2020) e00142
11

Attach 10 mm M5 screws, nuts, and M6 120 mm screws, nuts, and spacers (Fig. 21). After that, place the tray and the tray
lid, and connect the whole assemble to the frame base. If assembled like that, the tray will become locked by the spacers and
screws.
5.7. Drain pipe
Cut the 20 mm PVC pipe to a suitable size (from 120 to 140 mm long, 45°cuts; use a miter box and a saw to do the cuts)
and attach it to its holder (Fig. 22). Attach the holder to the front of the frame (Fig. 22).
5.8. Preparing the control board
The first step is to upload the firmware to the board using a computer. Download the Arduino IDE (https://www.arduino.
cc/en/Main/Software) and the Marlin package with the appropriate version of the firmware (https://osf.io/2c6t5/). This is an
updated version of Marlin used for previous autosamplers [16,17,22], with minor modifications, including the ability to con-
trol servo motors. More details about this Marlin package are provided elsewhere [16,22]. Upload it to the board.
The second step is to make sure the board can be accessed by the tablet. Use an OTG USB adapter and a USB cable to con-
nect the board to the tablet. The app for Android (version 7.0) that was used here was Serial USB Terminal
(https://play.google.com/store/apps/details?id = de.kai_morich.serial_usb_terminal&hl = en). It automatically recognizes
FTDI, the driver controlling the MKS Gen 1.4 board. Serial USB Terminal will tell immediately if the connection was success-
ful (Fig. 23). Change the BAUD rate to 115,200 to ensure proper connection.
5.9. Assembling the valve
Cut the valve lever on both sides and drill 3 mm holes matching those on the servo lever (Fig. 24). It may be necessary to
drill bigger holes on the servo lever.
Fig. 12. Syringe attached to the E axis.
M.C. Carvalho HardwareX 8 (2020) e00142
12

Attach the valve lever and the servo lever to each other using M3 screws and nuts. Connect both to the valve, and also
connect the 1/8 in. connectors and tubes (Fig. 25). The connectors must be firmly attached to the valve to avoid leaks.
Connect the servo to the control board (Fig. 26), and stablish a connection using Serial USB Terminal. Then, type
M280P0S0 and send the command. This should turn the servo to position 0.
Disconnect the servo from the board. Now that the position is known, attach the servo to the valve (Fig. 27).
Wrap the servo with the servo motor mount and the connectors with the valve servo connectors (Fig. 28). Attach all parts
using M3 screws.
Using 10 mm M5 screws, attach the servo-valve to the back of the frame (Fig. 29).
Fig. 13. Linear carrier for the Z axis, including the needle connector and the needle.
Fig. 14. Z axis extrusion profile, motor, and leadscrew.
M.C. Carvalho HardwareX 8 (2020) e00142
13

5.10. Attaching the control board
Attach the control board spacers to the control board using M3 screws and hammer nuts. Then, attach the board to the
left-back leg of the frame (Fig. 30). The USB port should point upwards.
Connect the stepper motor drivers to the control board (Fig. 30). Notice their proper orientation, and that for axes X and Y,
the drivers are A4988, and for Z and E are Drv8825.
5.11. Connecting cables and tubes
Connect the 4-wire cables between stepper motors and proper control sockets in the board (Fig. 31). Connect the servo
motor to its proper controller (Fig. 26).
Connect the 1/8 in. tubes to the 3 mm tubes (Fig. 32). This connection is straightforward for the tubes used here, without
need for connectors. This may not be the case for tubes from different brands. The top connector on the valve is connected to
Fig. 15. Complete Z axis, including the linear carrier.
Fig. 16. Motor mount and support plates for the X axis.
M.C. Carvalho HardwareX 8 (2020) e00142
14

the syringe. The one at position 0 (determined at step 5.9) is connected to the needle, and the one at position 180 is con-
nected to the sampling tube.
To power the board, connect positive to positive and negative to negative between battery and board power input.
5.12. Placing inside a container
It is possible to place the setup inside a large enough box, so that the system can be left reasonably protected from the
elements. A hole is necessary for the drain, but the sampling tube may be thin enough to go between the lid and the box
(Fig. 33).
Fig. 17. X axis plates connected.
Fig. 18. X axis fully assembled. Notice that the extrusion profile must have threaded holes for M6 screws.
M.C. Carvalho HardwareX 8 (2020) e00142
15

Fig. 19. X axis connected to the Y axis plates.
Fig. 20. Z axis connected to X axis.
M.C. Carvalho HardwareX 8 (2020) e00142
16

6. Operation instructions
6.1. Motor control
Make sure the board is connected to all motors and the battery. Connect it to the tablet, and use Serial USB Terminal to
control it. All commands sent to the board are a variation of G-code [30] for the Marlin firmware, as in previous autosamplers
[22,16,17]. The first command must be M121, to allow for negative coordinates. Stepper motors are controlled using the
command G1. For example, to move the syringe on the X axis, you can type G1X2F500. The parameter F is the feed rate, that
is, the speed of the movement. For the X axis, this is speed can be 500 units, while for the Y axis it should be 250 units. For the
Z axis, 1000 units work well. For the E axis, 150 is a safe value (higher values mean higher speeds; for the E axis, slow speeds
are needed because of the force needed to push water). F values can change depending on how the machine is built. A signal
that the value is too high is when the motor skip steps. If this happens, F values should be reduced until no skipping happens
anymore.
The command M280 controls the servos, which actuate the valves. M280P0S180 turns the first valve to the 180°position.
M280P0S0 turns the valve to the 0°position.
Fig. 21. Assembling the sample tray.
M.C. Carvalho HardwareX 8 (2020) e00142
17

6.2. Determining important positions
It is possible to adjust the needle and plunger positions either manually or using the procedures explained in Section 6.1.
Before using the autosampler, it is important that the needle and the plunger are at known positions, because the hardware
configuration presented here does not have limit switches. One way to do this is to determine the origins, or ‘‘zero”,
positions.
For the needle, the zero position chosen here was at the right / back corner, and top of the Z axis (Fig. 34). For the plunger,
it was the topmost position on the E axis (Fig. 34). Notice that at this position the tip of the plunger does not touch the body
of the syringe, but it is at about the 5 ml mark. It is important that the plunger does not go all the way up to the top, because
if so a vacuum will be created and when pulling the plunger down, the movement will be inaccurate. There is no mechanism
in place to prevent this to happen, so the user is responsible for ensuring that.
Once the needle and plunger are at the zero positions, the command G92X0Y0Z0E0 should be sent to the board, so that
the software will adopt these positions as zero. In this configuration, X, Y and E axes should use only positive numbers for the
movements, and Z only negative ones. If a different control board is used, it is possible that the values will be inverted.
Test motor movement with small increments (for example, G1X1F500). It is possible to reset positions by using the G92
command. For example, if you need to restart the X axis movement, you can manually push the needle back to zero position
for the X axis, and type G92X0.
Fig. 22. Drain pipe.
M.C. Carvalho HardwareX 8 (2020) e00142
18

Once the zero positions are determined, the coordinates for important positions for automated operation (Table 1) can be
found by using the G1 commands. It is always good idea to return the needle and plunger to zero positions once the proce-
dure is finished. Notice that all coordinates are relative to the zero positions, and that their units are arbitrary. Also, the coor-
dinates are for the machine built following the instructions presented here. If the machine is built differently, even only
slightly, coordinates will differ. In other words, the coordinates presented here should be only seen as examples of a possible
configuration, and not as parameters to be achieved when building the machine.
Fig. 23. Serial USB Terminal showing successful connection with control board.
M.C. Carvalho HardwareX 8 (2020) e00142
19

6.3. Automated control
Once the important positions for sampling have been determined, a script can be prepared to automate the sampling pro-
cedure, which will include a sequence of actions performed by the autosampler (Fig. 35).
This sequence of actions (Fig. 35) must become autosampler movements, which happens in the following sequence: 1)
the actions are translated into a Macrodroid script; 2) the actions in the Macrodroid script are sent to Serial USB Terminal;
3) Serial USB Terminal sends the actions to the control board; 4) the control board sends the actions to the motors; 5) the
motors actuate the syringe, plunger, or valve, and the movements happen in the real world.
Fig. 24. Modifying the valve and servo levers.
Fig. 25. Valve connected to levers and tubes.
M.C. Carvalho HardwareX 8 (2020) e00142
20

For example, the first action (Fig. 35) is ‘‘Move needle to drain”. In the real world, this means that the needle will move to
coordinates X = 6.0 and Y = 18.0, and after that to Z = 100 (values in Table 1; notice that horizontal movements must
always precede vertical ones). The respective G-code commands are G1X6.0F500, G1Y18.0F250, and G1Z100F1000. These
are sent from Macrodroid to USB Serial Terminal through a combination of clicks and character inputs, which are some of
the actions that can be automated using Macrodroid. For example, to send the first command, G1X6.0F500, Macrodroid per-
forms the following sequence of actions (see Fig. 36 for reference): 1) click on the writing field at the pixel at position 362,
372; 2) paste the value G1X6.0F500; 3) click on the send button (pixel 780, 732).
In addition to the sampling procedures, in order to save power, the tablet – control board connection is cut off and re-
stablished after each sample collection. Also, with the same purpose, the tablet display is turned off after all procedures,
and kept off for most of the time. It turns back on automatically when the sampling sequence re-starts for the next sample.
The full Macrodroid code, with a detailed explanation, is available in Supplementary Information 4. When editing the
code, it is not necessary that the tablet is connected to the control board.
6.4. Placement
The autosampler as presented here demands that the following conditions are satisfied:
1. It cannot be placed more than 2.5 m above water surface.
2. The sampling tube should not exceed 4 m.
Fig. 26. Servo cable connected to the control board.
Fig. 27. Servo connected to the valve.
M.C. Carvalho HardwareX 8 (2020) e00142
21

3. The autosampler should be placed on a flat, stable, horizontal surface. This is necessary to ensure that the stepper motors
performing the horizontal movements (X and Y axes) will reach the correct positions without trouble.
4. As explained in Section 6.2, the plunger of the syringe should not touch the top of the syringe to avoid generating extra
forces for the stepper motor actuating the syringe.
5. Depending on the sampling procedure, the waste should be contained, and disposed of in accordance to local environ-
mental regulations.
7. Validation and characterization
7.1. Carryover
One of the main concerns for an autosampler using a syringe pump is the carryover between samples. This carryover was
evaluated by measuring the dissolved organic carbon (DOC, measured using a Shimadzu TOC-L CSH/CSN analyser) content of
pure (milli-Q, or MQ) water alternately with concentrated glucose solutions, both sampled using the syringe pump-tube-
needle apparatus in the autosampler. DOC was chosen to test the autosampler for two main reasons: 1) it can be reliably
sampled using the autosampler if samples are preserved with acid, which precludes the need for refrigeration; 2) DOC is
an important parameter in environmental and ecological investigations [31,32], being widely measured for a number of dif-
ferent purposes, ranging from understanding the metabolism of unicellular organisms like bacteria and phytoplankton [33–
35], to understanding the carbon cycle at the ecosystem scale [36–38].
Several different kinds of samples were measured for DOC (Fig. 37). Five types of MQ water were measured: 1) no contact
with the autosampler; 2) passed through the autosampler before other solutions were measured; 3) passed through the
autosampler after 3 rinses following the most concentrated glucose solution; 4) same as 3, but after 5 rinses; 5) 3 rinses fol-
lowing the less concentrated glucose solution. There was no statistical difference (One-way ANOVA, F = 0.65, p = 0.59, sig-
nificance level p = 0.05) between the DOC in these MQ waters, but the MQ water sampled after 3 rinses following the most
concentrated glucose solution clearly contained more DOC than these other MQ solutions.
These experiments demonstrated that, although there is carryover between samples (as indicated by the larger DOC in
MQ water sampled after 3 rinses following the most concentrated glucose solution), this carryover is insignificant if a suf-
Fig. 28. Attaching servo to valve.
M.C. Carvalho HardwareX 8 (2020) e00142
22

ficient number of rinses is performed. The glucose concentrations used here, immediately followed by MQ, simulate an
extreme change in DOC that would be unlikely in nature. For example, DOC concentrations in rivers range typically between
0.15 and 1.3 mM, which are up to 170 times smaller than that of the most diluted solution used here, and, more importantly,
do not change abruptly between these extremes [38]. Therefore, it seems safe to limit the number of rinses to three to avoid
carryover between two consecutive samples. If the syringe or sampling line is modified, a different number of rinses may be
necessary. For 3 rinsing cycles, each sampling cycle takes 11 min. For faster sampling frequency, less rinses would be nec-
essary, at the risk of carryover. As explained here, the three rinses recommended here prevent even extreme carryover. For
systems that are enough known to not vary their conditions widely, a single rinse may be enough, thus enabling faster sam-
pling frequency.
7.2. Field test
The autosampler was deployed close to a creek inside the Southern Cross University, Lismore, NSW, Australia, on a rainy
day (1.4 mm, according to the Bureau of Meteorology), sampling hourly from 8:00 AM to 7:00 AM (Fig. 38).
Fig. 29. Servo attached to the frame.
M.C. Carvalho HardwareX 8 (2020) e00142
23

Fig. 30. Board attached to the back of the frame.
Fig. 31. Cables between board and motors.
M.C. Carvalho HardwareX 8 (2020) e00142
24

Fig. 32. Tubes connected to valve. Detail shows the connection between 1/8
00
and 3 mm tubes.
Fig. 33. Autosampler inside a box.
M.C. Carvalho HardwareX 8 (2020) e00142
25

A drop of concentrated (85%) phosphoric acid was added to each vial prior deployment in order to preserve samples for
total organic carbon (TOC; that is, DOC plus particulate carbon) measurements, which were done with the same equipment
used for the carryover test (Section 7.1). It was observed that TOC decreased through the day, to increase again early morning
next day (Fig. 39). It is possible that the decrease in TOC was due to dilution by the increased water volume brought by the
rain, which was then subdued after the rain was over (most of the rain fell between 11:00 am and 11:00 pm that day).
This test is preliminary and precludes firm conclusions about TOC dynamics in the creek. However, it serves to illustrate
how the autosampler could be used in a field sampling campaign. In particular, it was satisfactory that the protective box
allowed the system to work under rain.
7.3. Battery life
The field test (Section 7.3) also allowed the evaluation of the batteries supplying power to the autosampler. The 12 V bat-
tery powering the motors started at 13.20 V, and was at 11.65 V at the end of the sampling. The tablet battery started at 100%
Fig. 34. Initial positions for X, Y, Z and E axes. Below: detail of the plunger at its top position.
Table 1
Important positions for automated control. Notice that coordinates will differ from the values shown here if the machine is built with different dimensions and
if different boards or stepper motor drivers than those employed here are used.
Position description Coordinates
Safe vertical position for needle Z=0
Start vial (right/back corner) X = 2.6; Y = 1.3
End vial (left/front corner) X = 10.2; Y = 14.3
Needle inside vial Z=100
Drain pipe X = 6.0; Y = 18.0
Needle inside drain pipe Z=100
Plunger at empty position E=0
Plunger at full position E=95
Plunger at sampling position (that is, with the syringe holding the volume
of water to be delivered to a vial, in this case, 30 ml)
E=95
Valve at sucking position 180°
Valve at delivering position 0°
M.C. Carvalho HardwareX 8 (2020) e00142
26

Fig. 35. Sequence of actions for water sampling.
M.C. Carvalho HardwareX 8 (2020) e00142
27

and finished at 54%. In order to save power, the tablet was configured to turn off its screen when not in use. It is necessary
that the waiting time before the screen is turned off is set to at least 2 min, because some instructions take longer than a
minute to complete. Also, the brightness of the tablet display was reduced to its minimum level.
Using similar conditions (same number of rinses, dimmed tablet display, etc), uniformly timed sampling at 15 min, 3 h,
and 1 day were attempted. It was found that all 24 samples were collected for 15 min (totalling almost 12 h of activity). For
3 h intervals, only 17 samples were collected, and then the Tablet battery was flat. Accordingly, for samples spaced by 1 day,
only 4 samples were collected, and then the tablet battery was flat before the next sample. These tests indicate that the tablet
battery is a limiting factor for sampling, and a different power supply arrangement is necessary if longer sampling campaigns
are desired.
7.4. Possible modifications
The system presented here is about as simple as one can get for the task of water sampling in the field. Many improve-
ments are possible if more demanding sampling is necessary.
An obvious modification is to increase the size of the sampler, so that more samples, or larger volumes in larger bottles,
can be collected. If so, probably a second stepper motor for the Y axis would be necessary.
Fig. 36. Serial USB Terminal interface with the click positions highlighted.
M.C. Carvalho HardwareX 8 (2020) e00142
28

Fig. 37. Carryover test for the autosampler. Samples were measured using a TOC analyser. MQ stands for milli Q, which is very pure filtered water. Notice
that the most concentrated glucose solution was diluted 10 times for DOC measurement because the undiluted solution would generate significant
carryover in the TOC analyser. This solution was not diluted when going through the autosampler. Rinses were done using MQ water. The number of
replications per measurement is indicated by n. Notice that the X axis is on logarithmic scale.
Fig. 38. Autosampler close to a small creek in the Southern Cross University campus.
Fig. 39. Total organic carbon (TOC) from a creek collected using the autosampler presented in this paper during a 23 h cycle.
M.C. Carvalho HardwareX 8 (2020) e00142
29

Faster sampling could probably be achieved if a stronger stepper motor were used to power the syringe.
The current system employs timing belts for the X and Y axes. This is a low-cost option, but it is vulnerable to misposi-
tioning if the machine is moved around (for example, on a boat). Leadscrews would eliminate this problem. Also, limit
switches could be useful to minimize sampling errors.
All modifications listed above would increase the power consumption by the autosampler. As such, a larger battery, or
maybe a portable powering system using solar cells, for example, would be necessary.
Finally, a significant limitation of the autosampler presented here is that it cannot keep samples refrigerated, which is a
requirement for some water analyses. For these cases, a more complex, costly and energy-demanding system would be nec-
essary, possibly involving the use of a portable fridge.
Human and animal rights
The work did not involve human or animal subjects.
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.
Acknowledgements
I am grateful to Prof. Bradley Eyre, Southern Cross University, for allowing access to analytical equipment.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ohx.2020.e00142.
References
[1] S.C. Chapra, Surface Water-quality Modeling, McGraw-Hill, New York, 1997.
[2] R.G. Wetzel, G.E. Likens, Limnological Analyses, second ed., Springer-Verlag, New York, 1991.
[3] W. Jo, Y. Hoashi, L.L.P. Aguilar, M. Postigo-Malaga, J.M. Garcia-Bravo, B.C. Min, A low-cost and small USV platform for water quality monitoring,
HardwareX 6 (2019), https://doi.org/10.1016/j.ohx.2019.e00076.
[4] D.F. Carlson, A. Fürsteling, L. Vesterled, M. Skovby, S.S. Pedersen, C. Melvad, S. Rysgaard, An affordable and portable autonomous surface vehicle with
obstacle avoidance for coastal ocean monitoring, HardwareX 5 (2019), https://doi.org/10.1016/j.ohx.2019.e00059.
[5] A. Green, Z. Forsman, R.J. Toonen, M.J. Donahue, CoralCam: a flexible, low-cost ecological monitoring platform, HardwareX 7 (2020), https://doi.org/
10.1016/j.ohx.2019.e00089.
[6] A. Purser, U. Hoge, J. Lemburg, Y. Bodur, E. Schiller, J. Ludszuweit, J. Greinert, S. Dreutter, B. Dorschel, F. Wenzhöfer, PlasPI marine cameras: open-
source, affordable camera systems for time series marine studies, HardwareX 7 (2020), https://doi.org/10.1016/j.ohx.2020.e00102.
[7] G.M. Spinelli, Z.L. Gottesman, J. Deenik, A low-cost Arduino-based datalogger with cellular modem and FTP communication for irrigation water use
monitoring to enable access to CropManage, HardwareX 6 (2019), https://doi.org/10.1016/j.ohx.2019.e00066.
[8] A.P. Hill, P. Price, J.L. Snaddon, P. Doncaster, A. Rogers, AudioMoth: a low-cost acoustic device for monitoring biodiversity and the environment,
HardwareX 6 (2019), https://doi.org/10.1016/j.ohx.2019.e00073.
[9] A. Price, An apparatus for personalized atmospheric and flight data collection aboard high altitude weather balloons. HardwareX, 6, e00077 (2019).
doi:0.1016/j.ohx.2019.e00077.
[10] A.M. Matheny, P. Marchetto, J. Powell, A. Rechner, J.-Y. Chuah, E. McCormick, S.A. Pierce, LEAF: logger for ecological and atmospheric factors,
HardwareX 6 (2019).
[11] X. Mouy, M. Black, K. Cox, J. Qualley, C. Mireault, S. Dosso, F. Juanes, FishCam: a low-cost open source autonomous camera for aquatic research,
HardwareX 8 (2020), https://doi.org/10.1016/j.ohx.2020.e00110.
[12] ISY, ProSample Portable Samplers. https://www.ysi.com/prosample, (2020).
[13] Thermo-Fisher, MAXX Automatic Water Samplers. https://www.thermofisher.com.au/show.aspx?page=/ContentAUS/Environmental-Industrial/
Environmental-Monitoring-Safety/Water-Monitoring-Treatment/Surface-Water-Sampling-And-Monitoring/Automatic-Samplers.html, (2020).
[14] ISY, Wastewater-Stormwater Sampler. https://www.ysi.com/WS755, (2020).
[15] NOAA, Sub-surface automated sampler. https://www.coral.noaa.gov/accrete/sas/, (2019).
[16] M.C. Carvalho, R.H. Murray, Osmar, the open source microsyringe autosampler, HardwareX 3 (2018) 10–38, https://doi.org/10.1016/j.ohx.2018.01.001.
[17] M.C. Carvalho, C.J. Sanders, C. Holloway, Auto-HPGe, an autosampler for gamma-ray spectroscopy using high-purity germanium (HPGe) detectors and
heavy shields, HardwareX 4 (2018), https://doi.org/10.1016/j.ohx.2018.e00040.
[18] V.E. Garcia, J. Liu, J.L. DeRisi, Low-cost touchscreen driven programmable dual syringe pump for life science applications, HardwareX 4 (2018), https://
doi.org/10.1016/j.ohx.2018.e00027.
[19] V. Klar, J.M. Pearce, P. Karki, P. Kuosmanen, Ystruder: open source multifunction extruder with sensing and monitoring capabilities, HardwareX6
(2019), https://doi.org/10.1016/j.ohx.2019.e00080.
[20] B. Wijnen, E.J. Hunt, G.C. Anzalone, J.M. Pearce, Open-source syringe pump library. PLoS_One (2014), https://doi.org/10.1371/journal.pone.0107216.
doi:10.1371/journal.pone.0107216.
[21] T. Campbell, J.F.X. Jones, Design and implementation of a low cost, modular, adaptable and open-source XYZ positioning system for neurophysiology,
HardwareX 7 (2020), https://doi.org/10.1016/j.ohx.2020.e00098.
[22] M.C. Carvalho, W. Eickhoff, M. Drexl, Open-source autosampler for elemental and isotopic analyses of solids, HardwareX 8 (2020), https://doi.org/
10.1016/j.ohx.2020.e00123.
[23] G.E. Lancioni, N.N. Singh, M.F. O’Reilly, J. Sigafoos, G. Alberti, V. Perilli, V. Chiarello, G. Grillo, C. Turi, A tablet-based program to enable people with
intellectual and other disabilities to access leisure activities and video calls, Disability Rehabil.: Assistive Technol. 15 (2020) 14–20, https://doi.org/
10.1080/17483107.2018.1508515.
M.C. Carvalho HardwareX 8 (2020) e00142
30

[24] M.C.B. Pache, A.B. Costa, S.R. Souza, L.H. Negri, Speakcode: uma ferramenta de acessibilidade para pessoas com deficiencia visual, Rev. Bras. Educacao
Profissional tecnol. 1 (2020), https://doi.org/10.15628/rbept.2020.7934.
[25] G.E. Lancioni, M.F. O’Reilly, J. Sigafoos, G. Alberti, F. Campodonico, V. Chiarello, Promoting occupational engagement and personal satisfaction in people
with neurodevelopmental disorders via a smartphone-based intervention, Adv. Neurodev. Disorders 3 (2019) 259–266, https://doi.org/10.1007/
s41252-019-00102-4.
[26] M.C. Carvalho, Integration of analytical instruments with computer sripting, J. Lab. Autom. 18 (2013) 328–333.
[27] M.C. Carvalho, Practical Laboratory Automation made easy with AutoIt, Wiley VCH (2016).
[28] H. Miyajima, A study on measuring street slopes with Arduino for barrier free travel, Eco-Engineering 30 (2018) 111–114, https://doi.org/
10.11450/seitaikogaku.30.111.
[29] A.I. Ali, S.Z. Partal, S. Kepke, H.P. Partal, ZigBee and LoRa based wireless sensors for smart environment and IoT applications, in: 2019 1st Global Power,
Energy and Communication Conference (GPECOM), 2019, pp. 19–23. doi:10.1109/GPECOM.2019.8778505.
[30] P. Smid, CNC Programming Handbook, Industrial Press Inc, New York, 2003.
[31] J.I. Hedges, Why dissolved organics matter, in: D.A. Hansell, C.A. Carlson (Eds.) Biogeochemistry of Marine Dissolved Organic Matter, Academic Press,
San Diego, 2002, pp. 1–33.
[32] S.E.G. Findlay, R.L. Sinsabaugh, Aquatic Ecosystems: Interactivity of Dissolved Organic Matter, Academic Press, Amsterdan, 2003.
[33] J. Cherrier, J.E. Bauer, Bacterial utilization of transient plankton-derived dissolved organic carbon and nitrogen inputs in surface ocean waters, Aquat.
Microb. Ecol. 35 (2004) 229–241.
[34] J. Cherrier, S. Valentine, B. Hamil, W.H. Jeffrey, J. Marra, Light-mediated release of dissolved organic carbon by phytoplankton, J. Mar. Syst. 147 (2015)
45–51.
[35] N. González, J.-P. Gattuso, J.J. Middelburg, Oxygen production and carbon fixation in oligotrophic coastal bays and the relationship with gross and net
primary production, Aquat. Microb. Ecol. 52 (2008) 119–130.
[36] D.T. Maher, B.D. Eyre, Carbon budgets for three autotrophic Australian estuaries: Implications for global estimates of the coastal air-water CO2 flux,
Global Biogeochem. Cycles 26 (2012) GB1032.
[37] M.C. Carvalho, K.G. Schulz, B.D. Eyre, Respiration of new and old carbon in the surface ocean: implications for estimates of global oceanic gross primary
productivity, Global Biogeochem. Cycles 31 (2017).
[38] N.S. Wells, D. Maher, P. Huang, D.V. Erler, P. Maxwell, M.R. Hipsey, B.D. Eyre, Land-use intensity alters both the source and fate of CO2 within eight sub-
tropical estuaries, Geochim. Cosmochim. Acta 268 (2020) 107–122, https://doi.org/10.1016/j.gca.2019.09.042.
M.C. Carvalho HardwareX 8 (2020) e00142
31