Molekul, Vol. 20. No. 2, July 2025: 226 – 235

Articles

MOLEKUL
eISSN: 2503-0310

https://doi.org/10.20884/1.jm.2025.20.2.9230

Design and Fabrication of a Microfluidic Chip Device for Colorimetric Quantitative Analysis
of Albumin in Urine Matrix Solution
Karmanto1*, Retno Aliyatul Fikroh2, Laili Nailul Muna3
Department of Chemistry, Sunan Kalijaga State Islamic University, Yogyakarta, 55281, Indonesia
Department of Chemistry Education, Sunan Kalijaga State Islamic University, Yogyakarta, 5528, Indonesia
3
Department of Biomedical Sciences, Sunan Kalijaga State Islamic University, Yogyakarta, 5528, Indonesia
1

2

*Corresponding author email: karmanto@uin-suka.ac.id
Received September 06, 2023; Accepted June 19, 2025; Available online July 20, 2025
ABSTRACT. The existence of albumin in urine indicates kidney impairment, referred to as albuminuria. Early detection of
kidney dysfunction can be done through the examination of urine albumin levels. This research aims to develop a microfluidic
chip device for quantitative analysis of albumin in urine matrix solution using colorimetry with a smartphone-based device.
The microfluidic chip device for albumin is tested and evaluated through functional tests and analysis of the quality of
measurement data generated. The functional tests of the colorimetric system include testing the hydrophobic properties of
the fluid on the material surface, fluid distribution pattern, color appearance, and leakage. The quality of measurement data
is evaluated based on the predictive ability of the partial least squares regression models generated by data measurement
for predicting the albumin levels accurately and precisely. A prototype of the microfluidic chip device for albumin was made
from transparent acrylic material, with a 3-dimensional vertical fluid flow design consisting of two main parts: the inlet section
and the colorimetric system for visual fluid observation. Based on the functional tests, it is found that the acrylic material in
the colorimetric system has a hydrophilic surface characteristic with a contact angle of 61.15 o. The fluid can flow and
distribute well within the colorimetric system without producing bubbles. The color appearance in the colorimetric system is
also visualized well without fluid leakage. On the other hand, RGB HSL data, as the digital image identity of the training and
test samples, can be used as predictor variables. The method validation demonstrated acceptable sensitivity, with an LOD
of 3.368 mg/dL and an LOQ of 11.825 mg/dL. Precision at 16 mg/dL showed good repeatability (RSD 0.561%, recovery
99.26%;) and reproducibility (RSD 2.147%, recovery 101.42%). These results confirm the method's reliability for albumin
quantification at the tested concentration level.
Keywords: Albumin, albuminuria, colorimetry, microfluidic chip, PLSR, urine analysis

INTRODUCTION
The presence of albumin compounds in urine can
serve as indicators of kidney function and overall
health (Ketha & Singh, 2016). Increased albumin level
in urine indicate kidney dysfunction (Aitekenov et al.,
2021). Albumin is a protein usually present in high
concentrations in human blood serum (>30 mg/dL),
but its presence in urine should not exceed the clinical
threshold value of 30 mg/dL. Albuminuria is the
presence of a particular concentration of albumin in
the urine. Albuminuria is classified into two categories:
microalbuminuria if the rate of albumin excretion into
urine reaches 30-300 mg/24 hours, and
macroalbuminuria if the rate of albumin excretion
exceeds 300 mg/24 hours (Ketha & Singh, 2016).
Routine monitoring of albumin and creatinine
levels in urine is essential for assessing kidney function
(Delanghe et al., 2017), especially for high-risk
individuals with a history of diabetes and
cardiovascular disease (Comper & Osicka, 2005).
Various instruments and methods have been

developed for quantitative analysis of urinary albumin,
such as spectrophotometric determination of albuminbased on the ternary complex-formation reaction
(Yamaguchi et al., 2005), UPLC-MS-based analysis
(Ma et al., 2020), liquid chromatography-tandem
mass spectrometry (LC-MS/MS) (Ketha & Singh,
2016), and liquid chromatography-isotope dilution
tandem mass spectrometry (Chen et al., 2021). While
these methods provide accurate and reliable
measurement results (Aitekenov et al., 2021), the
primary instrumentation is typically large, heavy, nonportable, expensive, and complex. it requires skilled
personnel, leading to relatively high analysis costs (
Geng et al., 2023).
Overtime, on-site testing, and point-of-care
diagnostics need simple, fast, accurate, and
affordable analysis methods and instruments. Pointof-care diagnostics refers to the ability to perform
diagnostic tests and obtain results quickly at the exact
location of the patient, such as in clinics, hospitals, or
direct care environments. It means there is no need to

226

Design and Fabrication of a Microfluidic Chip Device

Karmanto, et al.

send samples to external laboratories and wait a long
time to receive results. Point-of-care diagnostic tests
are designed to provide immediate and direct results,
enabling prompt medical decision-making and
intervention. This technology is often utilized in urgent
situations, remote areas, or locations with limited
access to laboratory
facilities.
Point-of-care
diagnostics encompasses a wide range of tests,
including blood, urine, or other biological sample
tests, and can be conducted using portable devices or
specially developed smartphone applications.
Researchers have developed various portable
auxiliary devices for albumin analysis in the context of
point-of-care diagnostics. Some device platforms
developed, such as 3D-printed microfluidic chips
(Chan et al., 2016), paper-based microfluidic chips
(de Oliveira et al., 2017), paper-plastic hybrid
microfluidic chips (Uddin et al., 2017), paper-based
devices for visual detection (Hiraoka et al., 2020),
microfluidic
colorimetric
detection
platform
(Laurenciano et al., 2021), tags for in situ sensing (Jia
et al., 2022), and MIP dual electrochemical sensors for
one-step albumin analysis (Wardani et al., 2023).
Digital image colorimetry on smartphone is one of
the simple, easy, and affordable analysis methods
(Fan et al., 2021). Smartphone-based colorimetric
sensors enable convenient and immediate on-site
testing and point-of-care diagnostics, serving urgent
needs beyond the confines of a laboratory (Geng et
al., 2023). Smartphones are versatile analytical tools
that enable data acquisition, processing, and ondevice result generation (Böck et al., 2020). Several
application platforms, such as ColorX (Šafranko et al.,
2019), Color Analysis Pro, Albumin Tester (Coskun et
al., 2013), Albumin Smart Test (Mathaweesansurn et
al., 2017), and PhotoMetrix® (Böck et al., 2020), are
reported for smartphone-based colorimetric analysis.
Despite these advances, many of these strategies
remain constrained by cumbersome sample holders,
limited integration of fluidic control, and a reliance on
discrete calibration strips.
In this work, we introduce an acrylic-based
microfluidic chip featuring a 3D vertical-flow

architecture that uniquely combines precise fluid
handling with a simplified fabrication workflow. Acrylic
fabrication affords robust channel definition, optical
clarity for colorimetric readout, and compatibility with
low-cost laser-cutting or CNC processes. Together,
these features constitute a portable, accurate, and
user-friendly for point-of-care monitoring of urinary
albumin, advancing beyond prior paper-based and
lateral-flow devices.
EXPERIMENTAL SECTION
The materials utilized in this research included the
bromocresol green albumin assay kit (Sigma Aldrich,
USA), Albumin standard solution 5 g/dL (Sigma
Aldrich, USA), distillated water (Ikapharmindo
Putramas, Indonesia), and morning spot urine
samples. The equipment used comprised a 100 μL
microfluidic chip device made of acrylic material, a 550 μL micropipette, a 25 ml volumetric pipette, a 100
ml measuring flask, a 500 μL PCR vial, a digital lux
meter (Benetech), a HiView digital microscope (made
in China), and a Samsung Galaxy A325F/DS
smartphone equipped with a quad-camera rear
system. The primary camera boasts a 64MP
resolution, an 8MP ultra-wide camera, a 5MP macro
camera, and a 2MP depth sensor.
Design and Fabrication of Microfluidic Devices
A prototype microfluidic chip device for albumin
was developed, featuring a three-dimensional design
with a vertical fluidic flow mechanism. The microfluidic
chip design comprises two primary compartments: a
fluid inlet channel and a colorimetric system chamber.
The inlet channel is 1 mm in diameter and 5 mm in
length. The colorimetric system chamber measures
9.24 mm in diameter and 1 mm in thickness and can
hold 100 μL volume of fluid (Figure 1). Design and
fabrication of the microfluidic chip device involved the
utilization of Corel draw software and a laser cutting
machine. Following fabrication, the microfluidic chip
device underwent functional colorimetric tests,
including
assessments
of
fluid-material
hydrophobicity, fluid distribution patterns, color
visualization, and leakages.

Figure 1. Microfluidic Chip Design
227

Molekul, Vol. 20. No. 2, July 2025: 226 – 235
Preparation of Urine Solution
About 25 ml of urine was mixed with distilled water
in a measuring flask and diluted until the total volume
reached 100 ml. The mixture was then vigorously
shaken until a uniform solution was achieved.
Preparation of Albumin in Urine Matrix Solution.
The stock solution of the albumin in urine as matrix
solution was prepared by mixing 50 μL of the diluted
urine fluid with a certain amount of standard albumin
solution (5 g/dL) and distilled water to obtain albuminurine solutions with various concentration variations,
as detailed in Table 1.
Preparation of Training Sample
The training sample solutions for colorimetric
analysis (CS) were prepared by reacting five microliters
of albumin-urine solution from previous preparation
(AU) with 120 μL of bromocresol green (BCG) reagent
in Eppendorf tubes (Table 2) and allowed to stand at
room temperature for 5 minutes.
Preparation of Testing Sample
Testing samples were prepared with an albumin of
0.4 g/dL. The procedure for preparing the testing
sample was the same as the training sample, where it
was made by taking 5 μL of the albumin–urine stock

solution (AU-TS) and reacting it with 120 μL of the
BCG reagent solution, yielding an effective
concentration post-reagent addition of 16 mg/dL
(Table 3). The testing sample was left to stand for
5 minutes
before
conducting
colorimetric
measurements.
Preparation of Blank Solutions for LOD and LOQ
Determination
Seven matrix blank replicates were prepared by
mixing 5 µL of distilled water with 120 µL of BCG
reagent
and loading each mixture into the
microfluidic chip. The blanks were left to stand for
5 minutes
before
conducting
colorimetric
measurements under identical lighting and with the
same digital camera settings. From each image,
the mean RGB and HSL values within the region of
interest were extracted and submitted to the
pre-trained PLSR model to yield predicted blank
concentrations (𝑌𝐵 ). The overall blank mean (𝑌̂𝐵 ) and
standard deviation (𝑆𝐵 ) were calculated across all
replicates. The limits of detection (LOD) and
quantitation (LOQ) were then defined as 𝑌̂𝐵 + (3 × 𝑆𝐵 )
and 𝑌̂𝐵 + (10 × 𝑆𝐵 ), respectively, then interpreted
directly in units of mg/dL.

Table 1. Preparation of albumin in urine matrix solution (AU)
Albumin
stock solution
5 g/dL (ml)
100
80
40
12
2
0
20

ID
AU-1
AU-2
AU-3
AU-4
AU-5
Blank
AU-TS

Distilled
Water
(ml)
100
120
160
188
198
200
180

Urine
(ml)
50
50
50
50
50
50
50

Total
Volume
(ml)
250
250
250
250
250
250
250

[Albumin-Urine]
(g/dL)
2
1.6
0.8
0.24
0.04
0.00
0.4

Table 2. Preparation of training sample solution for colorimetric analysis
[Albumin-Urin]
(g/dL)
AU-1: 2
AU-2: 1.6
AU-3: 0.8
AU-4: 0.24
AU-6: 0.04
Blank: 0.00

ID
CS-1
CS-2
CS-3
CS-4
CS-5
Blank

Vol. [Albumin-Urin]
(ml)
5
5
5
5
5
0

*Effective concentration post-reagent addition

Vol. BCG
(ml)
120
120
120
120
120
120

[Albumin-Urin]*
(mg/dL)
80
64
32
9.6
1.6
0

Table 3. Preparation of testing sample
ID
TS-TR
TS-PR-1
TS-PR-2
TS-PR-3

AU-TS: 0.4 g/dL
(ml)
5
5
5
5

BCG
(ml)
120
120
120
120

*Effective concentration post-reagent addition
228

[Albumin-Urin]*
(mg/dL)
16
16
16
16

Design and Fabrication of a Microfluidic Chip Device

Karmanto, et al.

Colorimetric Analysis and Digital Image Processing
The colorimetric analysis was performed by
injecting 100 μL of the training sample and testing
sample solutions into a colorimetry system that
utilizes a microvolume chip device. The color
appearance observed on the microfluidic chip
device was recorded as a photograph using a
smartphone
camera under ambient lighting
conditions of 109.5 lux. The captured image was
subsequently processed and analyzed using Color
Analysis Pro software version 6 (Leizersoft) for
extracting RGB and HSL parameter values.
To control illumination‐induced variability in
image‐based measurements, the testing sample (TSTR) was photographed in triplicate under identical
smartphone camera and lighting conditions. These
technical replicates were employed to assess
repeatability of the RGB–HSL data extraction
procedure. Subsequently, to evaluate the variability
associated
with
sample
preparation,
three
independent aliquots of the 0.40 g/dL albumin sample
solution were prepared separately, and each was
photographed under identical lighting conditions
using the same camera device. These sample
preparative replicates (TS-PR) reflect intermediate
precision (within‐laboratory reproducibility), capturing
variability arising from pipetting, mixing, and reagent
addition.

was fabricated from acrylic materials, measures
38×12×3 mm in overall dimensions, and consists of
an inlet channel and a color-visualizing chamber. The
carefully designed microvolume inlet channel, with a
1 mm diameter and 5 mm length, prevents sudden
liquid spillage from the microfluidic chip device. It is
due to the significant atmospheric pressure difference
between the external and internal compartments. On
the other hand, the colorimetry visualizing chamber,
with a 9.24 mm in diameter and 1 mm thickness,
provides a larger and more spacious area for
convenient visual color observation and image
capture, even with microvolume-sized samples (see
Figure 2).

Building Predictive Models and Evaluating the Quality
of Measurement Data
The predictive model for analyzing albumin levels
in urine matrix solutions was performed using a
chemometric approach, i.e., Partial Least Squares
Regression (PLSR). This process was conducted using
Minitab 18 software (Minitab Inc., USA). The obtained
predictive model was evaluated for accuracy (%
recovery) and precision (RSD).
RESULTS AND DISCUSSION
Designing and fabricating a microfluidic chip device
The research focused on developing a microfluidic
chip device that serves as an assisting tool for
colorimetric quantitative albumin analysis. The device

Functional Evaluation of Colorimetry in a Microfluidic
Chip
The microfluidic chip's performance was evaluated
using four criteria: the surface hydrophobicity
properties of the material, fluid dispersion, color
appearance, and device leakage. As an assistive
colorimetric analysis tool, it is desirable to have a
surface material with hydrophilic characteristics that
align with the properties of the albumin-urine solution.
Additionally, the device should exhibit homogenous
fluid dispersion without air bubbles, enable observable
and easily distinguishable color appearance, and be
free from leakage.

Surface hydrophobicity properties of the microfluidic
chip material

The inclusion of carbonyl groups in the structure of
polymethyl methacrylate allows the acrylic material
utilized in microfluidic chip fabrication to possess
hydrophilic
surface
properties.
The
surface
hydrophobicity testing results confirm it (see Figure 3).
The acrylic material has a hydrophilic surface
characteristic with a contact angle of 61.15o. This
characteristic offers an advantage in which sample
solution droplets placed in the color visualization
compartment are securely held within the
microvolume space, preventing undesired spills or
leaks from the compartment area, even when the inlet
channel is positioned horizontally or upside down (see
Figure 6).

Figure 2: Prototype of a microfluidic chip device
229

Molekul, Vol. 20. No. 2, July 2025: 226 – 235

Figure 3. Prototype Surface hydrophobicity properties of the acrylic

Figure 4. Dispersion of the fluid in a microfluidic chip

Dispersion of the fluid in a microfluidic chip

Achieving a homogeneous fluid dispersion within
the microvolume space is crucial for obtaining an even
and consistent color appearance. Trapped air bubbles
within the color visualization chamber typically cause
non-uniform fluid dispersion. Removing these air
bubbles becomes challenging once they are formed
and trapped within the compartment because of the
fluid surface tension and the high atmospheric
pressure. Based on the observed results of the fluid
dispersion testing conducted in the microfluidic chip
device, it is evident that the device effectively retains
and evenly disperses the fluid, resulting in a
homogeneous and bubble-free fluid distribution
(Figure 4).

Color appearance of the fluid in a microfluidic chip

The color appearance of the sample solution in the
colorimetry visualization chamber is crucial and
influential for achieving successful colorimetric
analysis. The BCG reagent exhibits an initial yelloworange color, transitioning to a yellowish-green and
ultimately a bluish green (cyan) shade with increasing
albumin concentration (Figure 6). These color
variations serve as a visual indicator of the albumin
concentration in the analyzed sample solution. Despite
the small volume capacity, the color differences in the
sample fluid within the microfluidic chip device can be
easily observed.

Leakage testing of the microfluidic chip device

After assembling and connecting the acrylic chip
pieces, a leakage test is conducted to ensure there is
no fluid leakage outside the fluid flow system. Leakage
in microfluidic chip devices can cause changes in
volume and color stability. The leakage test involves
injecting a colored solution into the color visualization
chamber in a microfluidic chip device and allowing it
to remain for 12 hours for observation. There is no
leakage observed during the 12-hour observation
period. Additionally, the fluid does not easily spill even
when the chip is placed horizontally (Figure 6).
Colorimetric Analysis

Image processing

The colorimetric analysis involves capturing digital
images of both the calibration and test samples using
a smartphone. These digital images are then
processed using specialized software called Color
Analysis Pro. The software extracts various data,
including color group information, percentage of color
groups, specific color names, RGB color intensity, and
HSL values. An example of this analysis can be seen in
Figure 8. Table 4 presents the comprehensive
digital-image analysis results for the test samples,
including technical replicate test samples (to assess
repeatability) and preparative replicate test samples
(to assess reproducibility).

Figure 5. Color appearance of the fluid in a microfluidic chip
230

Design and Fabrication of a Microfluidic Chip Device

Karmanto, et al.

Figure 6. Leakage test of the microfluidic chip device

Predictive model development and colorimetric
albumin measurement

A Partial Least Squares Regression (PLSR) model
was constructed to predict albumin concentration
based on six colorimetric variables extracted from
digital images: Red (R), Green (G), Blue (B), Hue (H),
Saturation (S), and Lightness (L). The results of the PLSR
analysis are summarized in Table 5. The model
utilized leave-one-out cross-validation for component
selection. Among four components evaluated, the twocomponent model was selected as optimal, as it
provided the lowest predicted residual sum of squares
(PRESS = 197.29) and the highest predicted Rsquared value (R²ₚᵣₑd = 0.9658). The cumulative
variance explained by the two components in the X
matrix was 97.42%, indicating that the model
effectively captured the underlying variation in the
predictors. The selected PLSR model demonstrated
excellent performance with a high coefficient of
determination for the response variable albumin (R² =
0.9935), reflecting a strong linear relationship
between the predictors and the measured albumin
concentrations. Analysis of variance (ANOVA)
supported the statistical validity of the model, with an
F-value of 229.31 and a highly significant p-value (p
= 0.001), confirming that the regression model
significantly explained the variation in albumin levels
compared to the residual error.
Standardized regression coefficients revealed the
relative influence of each color parameter. Blue (B)

and Hue (H) contributed positively to the albumin
prediction, with standardized coefficients of 0.171 and
0.233, respectively. In contrast, Lightness (L) and
Green (G) exhibited strong negative contributions
(standardized coefficients of -0.213 and -0.237,
respectively). The variable Saturation (S) showed a
negligible effect. These results suggest that changes in
blue and hue intensities are associated with higher
albumin levels, while increases in green and lightness
reduce the predicted concentration. Based on these
PLSR-derived coefficients, the full linear regression
equation is as follows:
[𝐴𝑙𝑏𝑢𝑚𝑖𝑛] = 119.859 + (−0.240 × 𝑅) + (−0.602 × 𝐺 )
+ (0.667 × 𝐵) + (0.163 × 𝐻)
+ (−0.054 × 𝑆) + (−1.492 × 𝐿)

The model’s predictive performance was further
evaluated through residual analysis. The predicted
albumin values closely matched the actual
measurements across all samples, with small residuals
and standardized residuals within acceptable limits.
The PLS response plot for albumin, constructed from
the leave-one-out cross-validated “Fits(pred)” values
(calculated response) versus the observed albumin
concentrations (actual response), provides a direct
visual assessment of the two-component PLSR model’s
predictive performance. In this scatter plot, each point
represents one of the six urine samples, plotted with its
true concentration on the x-axis and its cross-validated
prediction on the y-axis (Figure 9).

Figure 8: Digital image processing and measurement of RGB color intensity and HSL values using
color analysis pro software.
231

Molekul, Vol. 20. No. 2, July 2025: 226 – 235
Table 4: Results of digital image analysis for the training sample (CS), test sample (TS) and blank
ID
CS-1
CS-2
CS-3
CS-4
CS-5
Blank
TS-TR-1
TS-TR-2
TS-TR-3
TS-PR-1
TS-PR-2
TS-PR-3
Blank-1
Blank-2
Blank-3
Blank-4
Blank-5
Blank-6
Blank-7

[Albumin]
mg/dL
80
64
32
9.6
1.6
0
16
16
16
16
16
16
0
0
0
0
0
0
0

R

G

B

H

S

L

72
74
103
129
137
146
121
120
121
122
120
119
146
140
145
141
143
142
144

112
115
132
141
137
143
138
137
138
137
135
135
143
140
143
138
141
138
139

103
99
101
95
79
92
94
93
96
96
92
92
92
80
94
87
86
82
90

167
157
116
76
60
57
83
83
84
82
81
82
57
60
58
57
58
56
54

22
22
13
19
27
23
19
19
18
18
19
19
23
27
21
24
25
27
23

36
37
46
46
42
47
45
45
46
46
45
45
47
43
47
45
45
44
46

Table 5: Predictive model building using PLSR analysis
PLS Regression: [Albumin] versus R, G, B, H, S, L method
Cross-validation
Components to evaluate
Number of components evaluated
Number of components selected

Leave-one-out
Set
4
2

Analysis of variance for [Albumin]
Source
Regression
Residual Error
Total

DF
2
3
5

SS
5736.56
37.52
5774.08

MS
2868.28
12.51

F
229.31

P
0.001

Model selection and validation for [Albumin]
Components
1
2
3
4

X Variance
0.703848
0.974209

Error
45.8624
37.5241
23.8945
0.5889

R-Sq
0.992057
0.993501
0.995862
0.999898

Coefficients
Constant
R
G
B
H
S
L

[Albumin]
119.859
-0.240
-0.602
0.667
0.163
-0.054
-1.492

[Albumin] standardized
0.000000
-0.227581
-0.236713
0.171239
0.233092
-0.007512
-0.212632

232

PRESS
289.011
197.288
816.146
946.507

R-Sq (pred)
0.949947
0.965832
0.858653
0.836077

Design and Fabrication of a Microfluidic Chip Device

Karmanto, et al.

Fits and residuals for [Albumin]
Row
1

[Albumin]
80.0

Fits
76.0654

Res
3.93455

SRes
1.65848

Fits (pred)
70.5529

Res (pred)
9.44709

2

64.0

67.9912

-3.99123

-1.49585

71.0917

-7.09165

3

32.0

32.5055

-0.50551

-0.26881

33.8828

-1.88282

4

9.6

9.9959

-0.39595

-0.13422

10.1353

-0.53526

5

1.6

2.7399

-1.13990

-0.53219

-4.8930

6.49300

6

0.0

-2.0980

2.09803

0.74409

-3.4291

3.42911

Figure 9. Scatterplot of observed versus predicted albumin concentrations
Table 6. Analytical sensitivity: LOD and LOQ

ID
Blank-1
Blank-2
Blank-3
Blank-4
Blank-5
Blank-6
Blank-7

Calculated
Blank
Concentration
mg/dL
-1.978
-0.495
-0.133
1.827
-1.017
-0.581
0.579

Mean (𝑌̂𝐵 )
mg/dL

Stdev (𝑆𝐵 )

LOD
𝑌̂𝐵 + (3 × 𝑆𝐵 )

LOQ
𝑌̂𝐵 + (10 × 𝑆𝐵 )

-0.257

1.208

3.368 mg/dL

11.825 mg/dL

Table 7. Percent recovery and percent RSD of repeatability
ID

[Albumin]
Actual

TS-TR-1

16 mg/dL

TS-TR-2

16 mg/dL

TS-TR-3

16 mg/dL

[Albumin]
Calculated

mean

Stdev

% Recovery

RSD

15.804
mg/dL
15.979
mg/dL
15.863
mg/dL

15.882
mg/dL

0.089

99.26%

0.561%

233

Molekul, Vol. 20. No. 2, July 2025: 226 – 235
Table 8 Percent recovery and percent RSD of reproducibility
ID

[Albumin]
Actual

TS-PR-1

16 mg/dL

TS-PR-2

16 mg/dL

TS-PR-3

16 mg/dL

[Albumin]
Calculated
15.889
mg/dL
16.19 mg/dL
16.593
mg/dL

mean

Stdev

% Recovery

RSD

16.227
mg/dL

0.0349

101.42%

2.147%

Method validation was carried out based on the
guidelines established by AOAC International (2016),
focusing on the evaluation of sensitivity, precision, and
accuracy. Sensitivity was assessed by analyzing seven
blank replicates (Table 6). The calculated mean blank
concentration was –0.257 mg/dL, with a standard

deviation of 1.208 mg/dL. From these values, the limit
of detection (LOD) and limit of quantitation (LOQ)
were calculated as 3.368 mg/dL and 11.825 mg/dL,
respectively. These results indicate that the method
possesses sufficient sensitivity for detecting albumin
within the expected working range.

Method validation

reproducibility (RSD 2.147%, recovery 101.42%), well
within the AOAC-specified limits. These results confirm
that the assay is both accurate and precise under the
evaluated conditions, supporting its suitability for
routine quantitative analysis. Future work should
extend
recovery
evaluation
across
multiple
concentration levels to fully characterize trueness
throughout the method’s dynamic range.

Precision
parameters,
repeatability
and
reproducibility, were evaluated at a single
concentration level of 16 mg/dL. Repeatability
(Table 7) was determined using three independently
captured measurements of the same sample, resulting
in a mean calculated concentration of 15.882 mg/dL,
a relative standard deviation (RSD) of 0.561%, and a
recovery of 99.26%. These values are within the
generally accepted criteria for repeatability, namely
RSD ≤3.7% and recovery within 95–105%.
Reproducibility (Table 8) was evaluated using three
independently prepared samples of the same
concentration. This yielded a mean calculated
concentration of 16.227 mg/dL, an RSD of 2.147%,
and a recovery of 101.42%, meeting the AOAC
criterion of RSD ≤6% for reproducibility (Latimer Jr. &
Latimer Jr., 2023).
Accuracy, expressed as percent recovery, was
examined only at the 16 mg/dL level. While the
recovery results at this concentration are acceptable,
the limited concentration range assessed restricts the
generalizability of the accuracy evaluation. Therefore,
additional validation at multiple concentration levels is
recommended in future work to ensure method
trueness across the full analytical range.
CONCLUSIONS
In conclusion, the developed acrylic‐based
microfluidic chip featuring three‐dimensional vertical
fluid flow through dedicated inlet channels and
colorimetric compartments demonstrated excellent
performance in terms of surface hydrophobicity, fluid
distribution, colorimetric response, and leak
resistance. The proposed microfluidic colorimetric
assay demonstrates robust analytical performance in
accordance with AOAC International (2016)
validation criteria. The method exhibits adequate
sensitivity, with an LOD of 3.368 mg/dL and an LOQ
of 11.825 mg/dL, ensuring reliable detection and
quantitation of albumin within the target range.
Precision testing at 16 mg/dL yielded excellent
repeatability (RSD 0.561%, recovery 99.26%) and

ACKNOWLEDGMENTS
We express our sincere gratitude for the financial
support provided by the Applied Research Grant of
UIN Sunan Kalijaga, based on the Rector's Decree No.
127.3 of 2021. This research project would not have
been possible without generous funding, and we are
immensely thankful for the opportunity to conduct this
study.
REFFERENCE
Aitekenov, S., Gaipov, A., & Bukasov, R. (2021).
Review: Detection and quantification of proteins
in human urine. Talanta, 223, 121718.
https://doi.org/10.1016/j.talanta.2020.121718
Böck, F. C., Helfer, G. A., da Costa, A. Ben, Dessuy,
M. B., & Ferrão, M. F. (2020). PhotoMetrix and
colorimetric image analysis using smartphones.
Journal of Chemometrics, 34(12), 1–19.
https://doi.org/10.1002/cem.3251
Chan, H. N., Shu, Y., Xiong, B., Chen, Y., Chen, Y.,
Tian, Q., Michael, S. A., Shen, B., & Wu, H.
(2016). Simple, cost-effective 3D printed
microfluidic components for disposable, pointof-care colorimetric analysis. ACS Sensors,
1(3),
227–234.
https://doi.org/10.1021/
acssensors.5b00100
Chen, Y., Liu, H., Loh, T. P., Liu, Q., Teo, T. L., Lee, T.
K., & Sethi, S. K. (2021). Measurement of urine
albumin by liquid chromatography-isotope
dilution tandem mass spectrometry and its
application to value assignment of external
quality assessment samples and certification of
reference materials. Clinical Chemistry and
Laboratory Medicine, 59(4), 711–720.

234

Design and Fabrication of a Microfluidic Chip Device

Karmanto, et al.

https://doi.org/10.1515/cclm-2020-0969
Comper, W. D., & Osicka, T. M. (2005). Detection of
urinary albumin. Advances in Chronic Kidney
Disease, 12(2), 170–176. https://doi.org/10.
1053/j.ackd.2005.01.012
Coskun, A. F., Nagi, R., Sadeghi, K., Phillips, S., &
Ozcan, A. (2013). Albumin testing in urine
using a smart-phone. Lab on a Chip, 13(21),
4231–4238.
https://doi.org/10.1039/c3lc
50785h
de Oliveira, R. A. G., Camargo, F., Pesquero, N. C.,
& Faria, R. C. (2017). A simple method to
produce 2D and 3D microfluidic paper-based
analytical devices for clinical analysis. Analytica
Chimica Acta, 957, 40–46. https://doi.org/
10.1016/j.aca.2017.01.002
Delanghe, J. R., Himpe, J., De Cock, N., Delanghe,
S., De Herde, K., Stove, V., & Speeckaert, M. M.
(2017). Sensitive albuminuria analysis using
dye-binding based test strips. Clinica Chimica
Acta, 471(March), 107–112. https://doi.org/
10.1016/j.cca.2017.05.032
Fan, Y., Li, J., Guo, Y., Xie, L., & Zhang, G. (2021).
Digital image colorimetry on smartphone for
chemical analysis: A review. Measurement:

Molecular Biology, 1378, 31–36. https://doi.

Journal of the International Measurement
Confederation, 171(December 2020), 1088

29. https://doi.org/10.1016/j.measurement.
2020.108829
Geng, Z., Miao, Y., Zhang, G., Liang, X. (2023).
Colorimetric biosensor based on smartphone:
State-of-art. Sensors and Actuators A: Physical,
349(1 January 2023 114056). https://doi.org/
https://doi.org/10.1016/j.sna.2022.114056
Hiraoka, R., Kuwahara, K., Wen, Y. C., Yen, T. H.,
Hiruta, Y., Cheng, C. M., & Citterio, D. (2020).
Paper-based device for naked eye urinary
albumin/creatinine ratio evaluation. ACS
Sensors, 5(4), 1110–1118. https://doi.org/10.
1021/acssensors.0c00050
Jia, Y., Liu, G., Xu, G., Li, X., Shi, Z., Cheng, C., Xu,
D., Lu, Y., Liu, Q. (2022). Battery-free and
wireless tag for in situ sensing of urinary
albumin/creatinine ratio (ACR) for the
assessment of albuminuria. Sensors and
Actuators B: Chemical, 367(15 September
2022), 132050. https://doi.org/https://doi.
org/10.1016/j.snb.2022.132050
Ketha, H., & Singh, R. J. (2016). Quantitation of
albumin in urine by liquid chromatography
tandem mass spectrometry. Methods in

org/10.1007/978-1-4939-3182-8_4
Latimer Jr., G. W., & Latimer Jr., G. W. (Eds.). (2023).
AF-1Guidelines
for
standard
method
performance requirements. In Official Methods
of Analysis of AOAC International (p. 0). Oxford
University Press. https://doi.org/10.1093/
9780197610145.005.006
Laurenciano, C. J. D., Tseng, C. C., Chen, S. J., Lu, S.
Y., Tayo, L. L., & Fu, L. M. (2021). Microfluidic
colorimetric detection platform with sliding
hybrid PMMA/paper microchip for human urine
and blood sample analysis.
Talanta,
231(March), 122362. https://doi.org/10.
1016/j.talanta.2021.122362
Ma, T., Liu, T., Xie, P., Jiang, S., Yi, W., Dai, P., & Guo,
X. (2020). UPLC-MS-based urine nontargeted
metabolic profiling identifies dysregulation of
pantothenate and CoA biosynthesis pathway in
diabetic kidney disease. Life Sciences, 258(6),
118160. https://doi.org/10.1016/j.lfs.2020.
118160
Mathaweesansurn, A., Maneerat, N., & Choengchan,
N. (2017). A mobile phone-based analyzer for
quantitative determination of urinary albumin
using self-calibration approach. Sensors and
Actuators, B: Chemical, 242, 476–483.
https://doi.org/10.1016/j.snb.2016.11.057
Šafranko, S., Živković, P., Stanković, A., MedvidovićKosanović, M., Széchenyi, A., & Jokić, S. (2019).
Designing ColorX, image processing software
for colorimetric determination of concentration,
to facilitate students’ investigation of analytical
chemistry concepts using digital imaging
technology. Journal of Chemical Education,
96(9), 1928–1937. https://doi.org/10.1021/
acs.jchemed.8b00920
Uddin, M. J., Jin, G. J., & Shim, J. S. (2017). Paperplastic
hybrid
microfluidic
device
for
smartphone-based colorimetric analysis of
urine. Analytical Chemistry, 89(24), 13160–
13166.
https://doi.org/10.1021/acs.
analchem.7b02612
Yamaguchi, T., Amano, E., Kamino, S., Umehara, S.,
Yanaihara, C., & Fujita, Y. (2005).
Spectrophotometric determination of urinary
protein
with
o-sulfophenylfluorone-metal
complex. Analytical Sciences, 21(10), 1237–
1240.
https://doi.org/10.2116/analsci.21.
1237

235