pc5271AY2526wiki - User contributions [en]

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:31:57Z

Peng Jianxi: /* 7. Repeatability / precision */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===
The peak intensity is defined as:
<math display="block">
I_{\text{peak}} = \max_{\lambda}, I_{\text{clean}}(\lambda)
</math>
This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and measurement alignment than integrated intensity. It is therefore used here primarily as an auxiliary metric and as the reference level for FWHM computation.

=== 4. Integrated intensity ===
The integrated emission intensity is defined as:
<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda), d\lambda
</math>
where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis. These bounds are chosen to include the full CQD emission band while excluding spectral regions dominated by residual background or solvent artifacts, and the same bounds are applied consistently across all samples and batches. Integrated intensity is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===
The spectral width is quantified using the full width at half maximum:
<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>
where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum first reaches half of the peak intensity on the short- and long-wavelength sides, respectively. This definition assumes a unimodal emission band; samples exhibiting multiple peaks or strongly asymmetric lineshapes may require additional characterization. FWHM is used as the main spectral-uniformity metric. A smaller FWHM indicates a narrower and more uniform emission band, while a larger FWHM indicates broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===
To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:
<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>
where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal computed in a wavelength region where no CQD photoluminescence is expected. In this work, the baseline region is taken from the short-wavelength side of the spectrum, below <math>\lambda_1</math>, where the CQD emission contribution is negligible. A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===
To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are performed. Two levels of repeatability are relevant in this work: within-batch repeatability, which reflects the stability of the measurement system for a single sample, and between-batch repeatability, which reflects the reproducibility of the synthesis process across independently prepared samples.

For any metric <math>x</math>, the mean and standard deviation over <math>n</math> repeated measurements are computed as:
<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>
<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>
The relative standard deviation (RSD) is then defined as:
<math display="block">
\mathrm{RSD}(x)% = \frac{s_x}{\bar{x}} \times 100%
</math>
In this work, the most useful repeatability metrics are:

<math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
<math>\mathrm{RSD}(I_{\text{int}})</math>
<math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:31:20Z

Peng Jianxi: /* 6. Signal-to-background ratio (SBR) */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===
The peak intensity is defined as:
<math display="block">
I_{\text{peak}} = \max_{\lambda}, I_{\text{clean}}(\lambda)
</math>
This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and measurement alignment than integrated intensity. It is therefore used here primarily as an auxiliary metric and as the reference level for FWHM computation.

=== 4. Integrated intensity ===
The integrated emission intensity is defined as:
<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda), d\lambda
</math>
where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis. These bounds are chosen to include the full CQD emission band while excluding spectral regions dominated by residual background or solvent artifacts, and the same bounds are applied consistently across all samples and batches. Integrated intensity is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===
The spectral width is quantified using the full width at half maximum:
<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>
where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum first reaches half of the peak intensity on the short- and long-wavelength sides, respectively. This definition assumes a unimodal emission band; samples exhibiting multiple peaks or strongly asymmetric lineshapes may require additional characterization. FWHM is used as the main spectral-uniformity metric. A smaller FWHM indicates a narrower and more uniform emission band, while a larger FWHM indicates broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===
To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:
<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>
where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal computed in a wavelength region where no CQD photoluminescence is expected. In this work, the baseline region is taken from the short-wavelength side of the spectrum, below <math>\lambda_1</math>, where the CQD emission contribution is negligible. A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:31:07Z

Peng Jianxi: /* 5. Full width at half maximum (FWHM) */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===
The peak intensity is defined as:
<math display="block">
I_{\text{peak}} = \max_{\lambda}, I_{\text{clean}}(\lambda)
</math>
This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and measurement alignment than integrated intensity. It is therefore used here primarily as an auxiliary metric and as the reference level for FWHM computation.

=== 4. Integrated intensity ===
The integrated emission intensity is defined as:
<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda), d\lambda
</math>
where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis. These bounds are chosen to include the full CQD emission band while excluding spectral regions dominated by residual background or solvent artifacts, and the same bounds are applied consistently across all samples and batches. Integrated intensity is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===
The spectral width is quantified using the full width at half maximum:
<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>
where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum first reaches half of the peak intensity on the short- and long-wavelength sides, respectively. This definition assumes a unimodal emission band; samples exhibiting multiple peaks or strongly asymmetric lineshapes may require additional characterization. FWHM is used as the main spectral-uniformity metric. A smaller FWHM indicates a narrower and more uniform emission band, while a larger FWHM indicates broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:30:51Z

Peng Jianxi: /* 4. Integrated intensity */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===
The peak intensity is defined as:
<math display="block">
I_{\text{peak}} = \max_{\lambda}, I_{\text{clean}}(\lambda)
</math>
This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and measurement alignment than integrated intensity. It is therefore used here primarily as an auxiliary metric and as the reference level for FWHM computation.

=== 4. Integrated intensity ===
The integrated emission intensity is defined as:
<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda), d\lambda
</math>
where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis. These bounds are chosen to include the full CQD emission band while excluding spectral regions dominated by residual background or solvent artifacts, and the same bounds are applied consistently across all samples and batches. Integrated intensity is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:30:35Z

Peng Jianxi: /* 3. Peak intensity */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===
The peak intensity is defined as:
<math display="block">
I_{\text{peak}} = \max_{\lambda}, I_{\text{clean}}(\lambda)
</math>
This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and measurement alignment than integrated intensity. It is therefore used here primarily as an auxiliary metric and as the reference level for FWHM computation.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:30:18Z

Peng Jianxi: /* 2. Peak wavelength */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===
The peak wavelength is defined as:
<math display="block">
\lambda_{\text{peak}} = \operatorname{argmax}{\lambda}, I{\text{clean}}(\lambda)
</math>
This indicates the dominant emission position and serves as a spectral-color and emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis conditions and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-17T07:29:48Z

Peng Jianxi: /* 1. Cleaned emission spectrum */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Difusion.jpg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

We recorded the diffusion of fluorescence in solution as a video,check the link:https://youtu.be/5WzAT8vx9_w?si=M06QTWSEPd1Xp3Lu

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===
The primary processed output is the cleaned CQD photoluminescence spectrum:
<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha, I_{\text{water}}(\lambda)
</math>
where:

<math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
<math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
<math>\alpha</math> is the background scaling factor, determined by minimizing the residual signal in a spectral region outside the known CQD emission band where no photoluminescence is expected.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:38:05Z

Peng Jianxi: /* Conclusion */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material throughout the surrounding liquid. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.
Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations that were difficult to correct for.
Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed satisfactorily.
Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality in a meaningful way. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for any future iteration of this work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:37:31Z

Peng Jianxi: /* Measurement pipelines */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer entrance while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference and improve measurement repeatability. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it more efficiently onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.

Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations.

Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed.

Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for future work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:36:28Z

Peng Jianxi: /* Introduction */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert security markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source and observed with the naked eye.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing functional groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax back. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.

Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations.

Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed.

Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for future work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:33:23Z

Peng Jianxi: /* Conclusion */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, the first objective was met. The prepared solution exhibited clear fluorescence under UV excitation, and droplets added to the solution produced a visible diffusion of fluorescent material. These observations, together with the peak positions recorded in the emission spectra, strongly suggest that fluorescent CQD species were successfully produced.

Quantitative characterization proved to be the main limitation. The Ocean Optics fiber spectrometer produced emission signals that were too weak for reliable measurements. Even with a focusing lens and the maximum available integration time, the signal-to-noise ratio remained insufficient for consistent results. Repeated measurements under nominally identical conditions also showed variability after dilution, and the absence of a fully dark environment introduced additional background fluctuations.

Sample handling over the multi-week project timeline further affected data quality. Without dedicated storage conditions, evaporation and precipitation gradually altered the sample concentration and composition, reducing control over experimental variables. Together with the limited sensitivity of the available instrumentation, these factors prevented several of the planned follow-up analyses from being completed.

Despite these limitations, the project successfully demonstrated CQD synthesis and showed that the measurement workflow can be used to assess solution quality. More sensitive instrumentation and tighter sample storage conditions would be the most impactful improvements for future work.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:32:38Z

Peng Jianxi: /* Measurement pipelines */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

Photoluminescence was excited using a commercial UV LED, angled at approximately 37° relative to the sample. This angle positions the brightest emission region toward the spectrometer while reducing the amount of scattered excitation light entering the detector.

Emission spectra were recorded with an Ocean Optics ST03418 spectrometer. Because CQDs produce relatively weak photoluminescence, good optical collection is important for achieving an adequate signal level and a favorable signal-to-background ratio. Measurements were therefore carried out in a dark room to suppress ambient light interference. A collection lens was also placed between the sample and the detector to capture a larger solid angle of emitted light and direct it onto the detector aperture.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, we achieved the first objective: the prepared solution exhibited clear fluorescence under UV excitation, and we observed a distinct diffusion of fluorescence upon adding droplets into the solution. These macroscopic signatures, together with the peak position visible in our recorded spectra, strongly suggest that fluorescent carbon quantum dot species were successfully produced.

However, quantitative characterization proved to be the main limitation. Using the Ocean Optics fiber spectrometer, the emission intensity remained too weak for high quality measurements. Even after adding a focusing lens and increasing the integration time to its maximum, the signal-to-noise ratio was insufficient, and repeated measurements under identical conditions cannot give us a fully reliable result after dilution. The lack of a true dark environment further increased background fluctuations and reduced repeatability.

In addition, practical constraints in sample handling affected data quality over the multi-week timeline. We did not have dedicated storage conditions for high-stability sample preservation, and evaporation and precipitation gradually changed the effective concentration and composition, compromising control over experimental variables. Combined with limited instrumental sensitivity, these factors prevented several planned follow-up analyses.

Overall, although our conclusions are not sufficiently rigorous due to various limitations, the project successfully demonstrated fluorescent CQD synthesis and established that our experimental and measurement workflow can be used to assess the quality of the CQD solution.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-12T05:32:02Z

Peng Jianxi: /* Introduction */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be dispersed in water and formulated into coatings or inks. When illuminated with UV light, these coatings produce visible emission, making them suitable for applications such as exhibition displays, interactive graphics, and covert markings.
This project uses CQDs synthesized from citric acid and urea. This precursor system is low-cost, relatively safe, and compatible with standard teaching-lab equipment. It also reliably produces bright fluorescence, typically in the green spectral region, which can be excited with a simple UV source.
CQDs from this system are prepared by heating an aqueous solution of the two precursors, either hydrothermally or in a microwave. During heating, the precursors dehydrate and partially carbonize, forming small carbon-rich particles. Oxygen- and nitrogen-containing groups remain on the particle surface. These surface groups, together with structural defects, create localized electronic states that govern the fluorescence: UV light promotes electrons to higher energy levels, and visible light is emitted as they relax. The emission wavelength and intensity depend on both the size of the carbon core and the nature of these surface states.
Heating conditions also influence the formation of small fluorescent molecular byproducts alongside the dots themselves. As a result, two batches prepared under slightly different time, power, or temperature conditions can differ noticeably in emission intensity and spectral shape.
The goal of this project is therefore to build a repeatable fluorescence measurement setup — UV excitation with spectrometer readout — and use it to evaluate CQD quality through objective metrics such as emission intensity and spectral consistency. These metrics are then related back to controllable synthesis parameters. The following sections outline the sensing approach, define the experimental objectives and measurables, and describe the synthesis and measurement procedures in detail.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepared fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

We describe the process of CQD synthesis:

# 1.0 g of citric acid and 1.0 g of urea were weighed and placed in a test tube.
# The mixture was moistened with approximately 3 drops of deionized water.
# Then we heated the mixture in a household microwave oven at full power. During heating, the reaction mixture frothed and gradually darkened.
# The heating process was stopped when the mixture became dark brown and the frothing had mostly subsided, which typically occurred after 2–3 min.

We excited the photoluminescence of the CQD with a commercial UV LED laser pointed at an angle about 37° as it provides effective excitation of the CQD solution while positioning the brightest fluorescent region toward the spectrometer and reducing direct scattered excitation light entering the detector.

We measured the fluorescence with an Ocean Optics [ST03418] spectrometer device. Since carbon quantum dots typically produce relatively weak photoluminescence, only a small fraction of the emitted light would reach the spectrometer without optical collection, resulting in a lower signal level and poorer signal-to-background ratio. Therefore, we conducted the experiment in a dark room to minimize interference from ambient light. Moreover, a collection lens was used to increase the collection efficiency by gathering a larger solid angle of emitted light and directing it toward the detector.

In our setup, the lens (focal length=18mm) was placed approximately 3.3 cm from the emitting region. Based on the thin-lens equation,

<math display="block">
\frac{1}{f}=\frac{1}{u}+\frac{1}{v},
</math>

where we substitute <math>f=18</math>mm and <math>u=33</math>mm, we get <math>v\approx3.95</math>cm.

This positive value of <math>v</math> indicates that the lens forms a real image of the emitting region. In practice, this means the lens can collect fluorescence from the sample and focus it more effectively into the detection path. A short focal length such as 18 mm is appropriate because it allows strong light collection from a nearby source and is well suited to weak fluorescence measurements.
The collection can also be estimated from the solid angle. For a lens at distance <math>u</math>, the collection half-angle is approximately
<math display="block">
\theta\approx \tan^{-1}(\frac{D/2}{u}),
</math>
where <math>D</math> is the lens diameter.

[[File:Experiment.jpeg|500px|center]]
[[File:Setup_R.jpg|500px|center]]
[[File:Setup.jpg|450px|center]]

Additionally, the solution concentration had a significant impact on the measurement results. Usable data were obtained only after the solution had been diluted until it became completely colorless and the spectrometer integral time had been increased to its maximum(6s). A possible explanation is that an excessive concentration of fluorescent species in the solution may partially block the emitted light.

For each setup, we measure two sets of spectra:

1. the '''water spectrum''', representing solvent and setup background,

2. the '''solution spectrum''', containing both CQD emission and background contributions.

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artefacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

=== repeatability check ===
[[File:Raw_rep.pdf|750px|center]]
[[File:Raw_dil.pdf|750px|center]]

There was a 7-day interval between the measurements for the first and second graphs. Due to solvent evaporation, we re-diluted the solution.

[[File:Dataset1.pdf|400px|center]]
[[File:Dataset2.pdf|400px|center]]
he first dataset corresponds to the initially prepared solution, while the second dataset corresponds to the solution after one week of storage followed by re-dilution.

Both datasets show a CQD-related emission feature , indicating CQD photoluminescence is present before and after dilution. The undiluted peak position is stable around <math>\lambda_{\text{peak}} \approx 547.5\,\text{nm}</math> (std <math>\sim 3.0\,\text{nm}</math>), while the diluted mean remains near <math>\lambda_{\text{peak}} \approx 544.6\,\text{nm}</math> but with much larger scatter (std <math>\sim 31\,\text{nm}</math>).The undiluted sample exhibits a strong, repeatable peak <math>I_{\text{peak}} \approx 689\,\text{a.u.}</math> (RSD <math>\sim 11.6\%</math>). After dilution, the peak collapses to <math>I_{\text{peak}} \approx 32.5\,\text{a.u.}</math> (<math>\sim 21\times</math> lower) and repeatability worsens (RSD <math>\sim 45\%</math>). CQD emission remains, but the usable fluorescence signal approaches the background level, amplifying sensitivity to environment light noise.

In the undiluted case, <math>I_{\mathrm{int}} \approx 6.91\times10^{4}\,\mathrm{a.u.}\cdot\mathrm{nm}</math> with RSD <math>\sim 11.2\%</math>, indicating a robust total-emission metric. After dilution, <math>I_{\mathrm{int}} \approx 1.46\times10^{3}\,\mathrm{a.u.}\cdot\mathrm{nm}</math>(<math>\sim 47\times</math> lower) and RSD rises to <math>\sim 86\%</math>,at the same time,the diluted dataset reports an unrealistically narrow and highly unstable width: FWHM <math>\approx 7.0\,\text{nm}</math> (RSD <math>\sim 57\%</math>). Such behavior is characteristic of low-SNR artifacts, where noise spikes or local fluctuations are misidentified as the main emission peak.Although both datasets show some degree of repeatability, the re-mixed and re-diluted solution exhibits a clear degradation in CQD quality.

== Conclusion ==

On the synthesis side, we achieved the first objective: the prepared solution exhibited clear fluorescence under UV excitation, and we observed a distinct diffusion of fluorescence upon adding droplets into the solution. These macroscopic signatures, together with the peak position visible in our recorded spectra, strongly suggest that fluorescent carbon quantum dot species were successfully produced.

However, quantitative characterization proved to be the main limitation. Using the Ocean Optics fiber spectrometer, the emission intensity remained too weak for high quality measurements. Even after adding a focusing lens and increasing the integration time to its maximum, the signal-to-noise ratio was insufficient, and repeated measurements under identical conditions cannot give us a fully reliable result after dilution. The lack of a true dark environment further increased background fluctuations and reduced repeatability.

In addition, practical constraints in sample handling affected data quality over the multi-week timeline. We did not have dedicated storage conditions for high-stability sample preservation, and evaporation and precipitation gradually changed the effective concentration and composition, compromising control over experimental variables. Combined with limited instrumental sensitivity, these factors prevented several planned follow-up analyses.

Overall, although our conclusions are not sufficiently rigorous due to various limitations, the project successfully demonstrated fluorescent CQD synthesis and established that our experimental and measurement workflow can be used to assess the quality of the CQD solution.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-04-08T05:23:06Z

Peng Jianxi: /* Measurement pipelines */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be formulated into water-based coatings and inks, enabling UV-activated visuals for applications such as exhibition displays, interactive graphics, and covert markings.
In this project, we focus on CQDs synthesized from citric acid and urea because the system is low-cost, relatively safe, and easy to prepare with typical teaching-lab equipment, and it commonly yields bright visible fluorescence (often in the green region) that can be excited using a simple UV source.

Citric acid + urea CQDs are typically made by heating an aqueous precursor solution (e.g., hydrothermal or microwave). During heating, the molecules dehydrate and partially carbonize, forming tiny carbon-rich particles while leaving oxygen- and nitrogen-containing groups on the surface. These CQDs fluoresce because UV/blue light excites electrons to higher energy, and when the electrons relax, they emit visible light; the brightness and color depend strongly on the size/structure of the carbon core and, especially, on surface “states” created by surface groups and defects.
In this system, the heating conditions can also produce small fluorescent byproducts in addition to the dots, so two batches made with slightly different time/power/temperature can have noticeably different intensity and spectral shape.
Therefore, our goal is to build and use a repeatable fluorescence sensing setup (UV excitation + spectrometer readout) to quantitatively evaluate CQD “quality” using objective metrics such as fluorescence intensity and spectral consistency, and to relate these metrics back to controllable synthesis parameters.

In the sections that follow, we outline the sensing rationale, define our experimental goals and measurables, and then detail the synthesis and fluorescence-measurement methodology used to quantify CQD quality.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepare fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

For synthesis, we dissolve '''1 g citric acid''' and '''1 g urea''' in '''deionized water''' using standard laboratory glassware (beakers, volumetric flasks, and pipettes), then heat the solution in a '''microwave oven''' under controlled conditions until CQDs form. The specific steps are as follows:
1. Prepare the Reagents: Weigh approximately 1 gram of urea and 1 gram of citric acid, and add them both into a test tube.
2. Moisten the Mixture: Add about 3 drops of deionized water to the reaction mixture in the test tube to moisten it.
3, Microwave Heating: Place the test tube in a household microwave oven and heat it on full power.
4. Observe and Stop: During the heating process, the mixture will froth and gradually turn darker. Stop the heating once the mixture becomes dark brown and the frothing has mostly subsided. This usually takes about 2 to 3 minutes.

The optical setup uses:
* a commercial UV LED light pen as the excitation source,
* an Ocean Optics '''[insert specs]''' spectrometer, as the detector,
* '''[insert lens specs]''' for better light capturing
* deionized water (same as used in the sample) as the reference/background measurement.

Two spectra are measured under identical conditions:
# the '''water spectrum''', representing solvent and setup background,

# the '''solution spectrum''', containing both CQD emission and background contributions.

The cleaned CQD emission spectrum is obtained by subtracting the water reference from the solution signal:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

The excitation-leakage / Rayleigh-scattering region near the LED wavelength is excluded from analysis, and isolated detector spikes are removed before final metric extraction.

'''[INSERT IMAGES]'''

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artifacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

[[File:Experiment.jpeg|750px|center]]

=== Apparatus set up ===

[[File:Setup_R.jpg|750px|center]]
[[File:Setup.jpg|750px|center]]
=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-03-18T05:20:30Z

Peng Jianxi: /* Data-processing workflow */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be formulated into water-based coatings and inks, enabling UV-activated visuals for applications such as exhibition displays, interactive graphics, and covert markings.
In this project, we focus on CQDs synthesized from citric acid and urea because the system is low-cost, relatively safe, and easy to prepare with typical teaching-lab equipment, and it commonly yields bright visible fluorescence (often in the green region) that can be excited using a simple UV source.

Citric acid + urea CQDs are typically made by heating an aqueous precursor solution (e.g., hydrothermal or microwave). During heating, the molecules dehydrate and partially carbonize, forming tiny carbon-rich particles while leaving oxygen- and nitrogen-containing groups on the surface. These CQDs fluoresce because UV/blue light excites electrons to higher energy, and when the electrons relax, they emit visible light; the brightness and color depend strongly on the size/structure of the carbon core and, especially, on surface “states” created by surface groups and defects.
In this system, the heating conditions can also produce small fluorescent byproducts in addition to the dots, so two batches made with slightly different time/power/temperature can have noticeably different intensity and spectral shape.
Therefore, our goal is to build and use a repeatable fluorescence sensing setup (UV excitation + spectrometer readout) to quantitatively evaluate CQD “quality” using objective metrics such as fluorescence intensity and spectral consistency, and to relate these metrics back to controllable synthesis parameters.

In the sections that follow, we outline the sensing rationale, define our experimental goals and measurables, and then detail the synthesis and fluorescence-measurement methodology used to quantify CQD quality.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepare fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

For synthesis, we dissolve '''1 g citric acid''' and '''1 g urea''' in '''deionized water''' using standard laboratory glassware (beakers, volumetric flasks, and pipettes), then heat the solution in a '''microwave oven''' under controlled conditions until CQDs form.

The optical setup uses:
* a commercial UV LED light pen as the excitation source,
* an Ocean Optics '''[insert specs]''' spectrometer, as the detector,
* '''[insert lens specs]''' for better light capturing
* deionized water (same as used in the sample) as the reference/background measurement.

Two spectra are measured under identical conditions:
# the '''water spectrum''', representing solvent and setup background,

# the '''solution spectrum''', containing both CQD emission and background contributions.

The cleaned CQD emission spectrum is obtained by subtracting the water reference from the solution signal:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

The excitation-leakage / Rayleigh-scattering region near the LED wavelength is excluded from analysis, and isolated detector spikes are removed before final metric extraction.

'''[INSERT IMAGES]'''

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artifacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

[[File:Experiment.jpeg|750px|center]]

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.

File:Experiment.jpeg

2026-03-18T05:12:22Z

Peng Jianxi:

Fluorescence Sensor for Carbon Quantum Dots: Synthesis, Characterization, and Quality Control

2026-03-17T13:19:44Z

Peng Jianxi: /* Data-processing workflow */

== Introduction ==
Carbon quantum dots (CQDs) are fluorescent nanomaterials that can be formulated into water-based coatings and inks, enabling UV-activated visuals for applications such as exhibition displays, interactive graphics, and covert markings.
In this project, we focus on CQDs synthesized from citric acid and urea because the system is low-cost, relatively safe, and easy to prepare with typical teaching-lab equipment, and it commonly yields bright visible fluorescence (often in the green region) that can be excited using a simple UV source.

Citric acid + urea CQDs are typically made by heating an aqueous precursor solution (e.g., hydrothermal or microwave). During heating, the molecules dehydrate and partially carbonize, forming tiny carbon-rich particles while leaving oxygen- and nitrogen-containing groups on the surface. These CQDs fluoresce because UV/blue light excites electrons to higher energy, and when the electrons relax, they emit visible light; the brightness and color depend strongly on the size/structure of the carbon core and, especially, on surface “states” created by surface groups and defects.
In this system, the heating conditions can also produce small fluorescent byproducts in addition to the dots, so two batches made with slightly different time/power/temperature can have noticeably different intensity and spectral shape.
Therefore, our goal is to build and use a repeatable fluorescence sensing setup (UV excitation + spectrometer readout) to quantitatively evaluate CQD “quality” using objective metrics such as fluorescence intensity and spectral consistency, and to relate these metrics back to controllable synthesis parameters.

In the sections that follow, we outline the sensing rationale, define our experimental goals and measurables, and then detail the synthesis and fluorescence-measurement methodology used to quantify CQD quality.

== Report objectives and scope==

The objective of this work is to quantify the '''optical fluorescence quality''' of citric-acid/urea carbon quantum dots (CQDs) using a repeatable optical setup based on '''UV LED excitation''' and '''spectrometer detection'''. The goal is to generate metrics that allow comparison of CQD batches produced under different synthesis conditions.

The analysis is designed to:

# isolate true photoluminescence (PL) from optical artifacts,
# extract robust spectral quality metrics,
# assess whether the emission is relatively uniform or heterogeneous,
# evaluate whether observed differences are real or caused by measurement noise or drift.

In this project, “quality” refers to optical fluorescence quality, not full chemical or structural quality.

A higher-quality CQD solution is defined as one that shows:
* stronger fluorescence under fixed measurement conditions,
* a cleaner and more stable emission spectrum,
* a narrower and more uniform emission band,
* better repeatability across repeated measurements.

Thus, CQD quality is evaluated through four dimensions:

<math display="block">
\text{Quality}\sim\text{brightness} + \text{spectral purity} + \text{spectral uniformity} + \text{measurement repeatability}
</math>

== Measurement pipelines ==

We prepare fluorescent carbon quantum dots (CQDs) by heating an aqueous mixture of citric acid and urea, which promotes dehydration and condensation of the precursors, followed by partial carbonization to form nanoscale carbonaceous particles. Citric acid serves as the carbon source, while urea introduces nitrogen-containing surface functional groups that improve water dispersibility and typically increase fluorescence intensity.

For synthesis, we dissolve '''1 g citric acid''' and '''1 g urea''' in '''deionized water''' using standard laboratory glassware (beakers, volumetric flasks, and pipettes), then heat the solution in a '''microwave oven''' under controlled conditions until CQDs form.

The optical setup uses:
* a commercial UV LED light pen as the excitation source,
* an Ocean Optics '''[insert specs]''' spectrometer, as the detector,
* '''[insert lens specs]''' for better light capturing
* deionized water (same as used in the sample) as the reference/background measurement.

Two spectra are measured under identical conditions:
# the '''water spectrum''', representing solvent and setup background,

# the '''solution spectrum''', containing both CQD emission and background contributions.

The cleaned CQD emission spectrum is obtained by subtracting the water reference from the solution signal:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

The excitation-leakage / Rayleigh-scattering region near the LED wavelength is excluded from analysis, and isolated detector spikes are removed before final metric extraction.

'''[INSERT IMAGES]'''

== Data-processing workflow ==
To ensure the reported spectra accurately reflect the intrinsic photoluminescence (PL) of the Carbon Quantum Dots (CQDs), the following post-acquisition processing protocol was implemented:

# Acquire raw spectra of water and CQD solution under the identical optoelectronic configurations conditions using the '''OceanView''' software.
# Align the spectra in wavelength.
# Subtract the solvent blank from the solution spectrum.
# Exclude the wavelength region proximal to the excitation wavelength to remove excitation leakage and Rayleigh scattering.
# Remove isolated detector spikes and nonphysical subtraction artifacts.
# Use the processed spectrum for metric extraction.

Subsequently, we report the following quantities for each batch.

=== 1. Cleaned emission spectrum ===

The primary processed output is the cleaned CQD photoluminescence spectrum:

<math display="block">
I_{\text{clean}}(\lambda) = I_{\text{solution}}(\lambda) - \alpha I_{\text{water}}(\lambda)
</math>

where:
* <math>I_{\text{solution}}(\lambda)</math> is the measured CQD solution spectrum,
* <math>I_{\text{water}}(\lambda)</math> is the water reference spectrum,
* <math>\alpha</math> is the background scaling factor.

This cleaned spectrum is used for all subsequent metric extraction.

=== 2. Peak wavelength ===

The peak wavelength is defined as:

<math display="block">
\lambda_{\text{peak}} = \text{argmax}_{\lambda} I_{\text{clean}}(\lambda)
</math>

This indicates the dominant emission position and serves as a spectral-color / emissive-state indicator. Stable values of <math>\lambda_{\text{peak}}</math> across repeated measurements or across batches suggest more consistent synthesis and more reproducible emissive behavior.

=== 3. Peak intensity ===

The peak intensity is defined as:

<math display="block">
I_{\text{peak}} = \max_{\lambda} I_{\text{clean}}(\lambda)
</math>

This gives the maximum signal level in the cleaned spectrum. It is useful as a simple brightness indicator, but it is more sensitive to noise and alignment than integrated intensity.

=== 4. Integrated intensity ===

The integrated emission intensity is defined as:

<math display="block">
I_{\text{int}} = \int_{\lambda_1}^{\lambda_2} I_{\text{clean}}(\lambda)\, d\lambda
</math>

where <math>\lambda_1</math> and <math>\lambda_2</math> define the valid emission range used for analysis.

This is the main brightness metric used in this work because it captures the total emitted signal and is more robust than peak intensity for comparing different batches under fixed measurement conditions.

=== 5. Full width at half maximum (FWHM) ===

The spectral width is quantified using the full width at half maximum:

<math display="block">
\mathrm{FWHM} = \lambda_{\text{right}} - \lambda_{\text{left}}
</math>

where <math>\lambda_{\text{left}}</math> and <math>\lambda_{\text{right}}</math> are the wavelengths at which the cleaned spectrum reaches half of the peak intensity.

FWHM is used as the main spectral-uniformity metric. A smaller FWHM suggests a narrower and more uniform emission band, while a larger FWHM suggests broader emission and greater heterogeneity of emissive states.

=== 6. Signal-to-background ratio (SBR) ===

To evaluate measurement quality and confidence in the extracted signal, the signal-to-background ratio is defined as:

<math display="block">
\mathrm{SBR} = \frac{I_{\text{peak}}}{\sigma_{\text{baseline}}}
</math>

where <math>\sigma_{\text{baseline}}</math> is the standard deviation of the cleaned signal in a wavelength region where no real CQD photoluminescence is expected.

A larger SBR indicates that the detected fluorescence is strong relative to the residual background noise, and therefore more reliable.

=== 7. Repeatability / precision ===

To assess whether differences between samples are real rather than caused by setup drift or noise, repeated measurements are used. For any metric <math>x</math>, the mean and standard deviation are computed as:

<math display="block">
\bar{x} = \frac{1}{n}\sum_{k=1}^{n} x_k
</math>

<math display="block">
s_x = \sqrt{\frac{1}{n-1}\sum_{k=1}^{n}(x_k - \bar{x})^2}
</math>

The relative standard deviation (RSD) is then defined as:

<math display="block">
\mathrm{RSD}(x)\% = \frac{s_x}{\bar{x}} \times 100\%
</math>

In this work, the most useful repeatability metrics are:
* <math>\mathrm{RSD}(\lambda_{\text{peak}})</math>
* <math>\mathrm{RSD}(I_{\text{int}})</math>
* <math>\mathrm{RSD}(\mathrm{FWHM})</math>

A lower RSD indicates better measurement precision and higher confidence in batch-to-batch comparisons.

=== 8. Summary of key quality metrics ===

The primary quality metrics used in this work are:
* <math>I_{\text{int}}</math> as the main brightness metric,
* <math>\lambda_{\text{peak}}</math> as the dominant emission-position metric,
* <math>\mathrm{FWHM}</math> as the main spectral-uniformity / heterogeneity metric,
* <math>\mathrm{SBR}</math> as a measurement-confidence metric,
* repeatability metrics based on <math>\mathrm{RSD}</math> to quantify precision.

Together, these quantities provide a practical framework for comparing the optical fluorescence quality of CQD batches produced under different synthesis conditions.

== Results and discussions ==
[[File:cqd_water_solution_compare.pdf|750px|center]]

The background-subtracted spectrum shows a clear broad positive emission band, confirming that the CQD solution produces genuine photoluminescence rather than only solvent or instrumental background. The dominant emission peak is located at <math>\lambda_{\mathrm{peak}} = 548.28\ \mathrm{nm}</math>, placing the fluorescence in the green region of the visible spectrum. The extracted peak intensity is <math>I_{\mathrm{peak}} = 717.72\ \mathrm{a.u.}</math>, while the integrated intensity is <math>I_{\mathrm{int}} = 73457.29\ \mathrm{a.u.\ nm}</math>, indicating that the sample is clearly fluorescent under the present measurement conditions.

The emission width is relatively broad, with <math>\mathrm{FWHM} = 91.73\ \mathrm{nm}</math>. This suggests that the fluorescence is not produced by a highly uniform single emissive state, but more likely by a distribution of emissive environments or overlapping emissive centers. In other words, the sample shows strong fluorescence but only moderate spectral uniformity.

Several narrow spikes and sharp dips remain in the cleaned raw trace. These are much narrower than the main emission band and are therefore attributed to detector or subtraction artifacts rather than intrinsic CQD features. Overall, the result indicates a fluorescent green-emitting CQD sample with noticeable spectral heterogeneity.