Development and performance testing of a high-sensitivity pulsed flame photometric detector

doi:10.3724/SP.J.1123.2026.03027

Abstract

Abstract:

The conventional flame photometric detector （FPD） is widely used for sulfur-selective gas chromatography. However， the FPD is sensitive to hydrocarbon backgrounds. Its sulfur response also depends strongly on flame stability. These factors limit stable trace sulfur determination in complex matrices. A pulsed flame photometric detector （PFPD） provides a time-resolved approach to this problem. Sulfur emission appears later than most hydrocarbon background emissions after pulsed combustion. A suitable acquisition window can therefore improve sulfur selectivity. In this study， a high-sensitivity single-channel PFPD was designed and tested. The aim was to establish reproducible operating conditions for trace sulfur analysis. The detector consisted of five functional modules： the combustion and ignition module， the gas supply module， the optical collection and detection module， the temperature-control module， and the mechanical support and sealing module. The combustion zone consisted of a vertically mounted quartz chamber. The chamber had an outer diameter of 4 mm. Its inner diameter was 2 mm. Its total length was 17.5 mm. The effective combustion length was 12 mm. The calculated chamber volume was 37.68 mm³. Numerical simulation was used to examine flame propagation and self-extinction. A 12 mm effective length maintained a stable finger-shaped flame front. The flame was self-terminated near the lower limiting orifice at about 1.97 ms. The hydroxyl radical mass fraction decreased close to baseline within 5 ms， whereas the temperature decayed more gradually and approached its initial level at approximately 30 ms. These results supported the chamber design for periodic pulsed combustion. The optical unit used a quartz light guide and a BG-12 band-pass filter. The filter matched the sulfur emission band near 394 nm. A photomultiplier tube （PMT） recorded the time-resolved emission signal. Two electronic pneumatic control （EPC） units were used for gas regulation. One unit controlled hydrogen and air for the ignition chamber. The other unit controlled hydrogen and air for the combustion chamber. This design separated ignition conditions from analytical combustion conditions. It also reduced uncertainty caused by coupled gas adjustment. The optimization started with the ignition chamber. Then the combustion chamber was optimized under the fixed ignition baseline. Finally， detector temperature and PMT voltage were evaluated. Peak area， sensitivity， and signal-to-noise ratio （SNR） were used as indicators. Gas chromatographic tests used a sulfur-specific quartz capillary column. Nitrogen was used as the carrier gas. The gate delay was 6 ms. The gate width was 18 ms. The trigger level was 500 mV. A stable pulsed flame was obtained at an ignition-chamber hydrogen flow of 8.0 mL/min. The corresponding ignition-chamber air flow was 19.0 mL/min. Under this ignition baseline， the combustion chamber was further adjusted. A higher sulfur response was obtained at 5.0 mL/min hydrogen. The corresponding air flow was 3.7 mL/min. Detector temperature also affected the delayed sulfur emission. The response was low at 110-130 ℃. At about 150 ℃， the delayed sulfur signal was higher. Peak area and sulfur sensitivity were also higher at this temperature. Higher temperatures from 170 to 250 ℃ reduced the peak area and sensitivity. PMT voltage affected signal amplitude， SNR， and waveform fidelity. Increasing the voltage from 590 to 710 V improved the SNR. At 730 V， the delayed emission plateau showed peak clipping. This behavior indicated saturation or nonlinear response in the detection chain. Dynamic dilutions of a 10 μmol/mol hydrogen sulfide （H₂S） standard gas were used to evaluate the linear range. A 1.1 μmol/mol H₂S standard gas was used for the quantitative repeatability test and， after 30-fold dynamic dilution， for determination of the sulfur detection limit. The quantitative repeatability， evaluated from the peak areas of seven consecutive injections， was 0.37%. The sulfur detection limit was 4.1×10⁻¹³ g/s， and the linear range spanned approximately two orders of magnitude. The developed PFPD showed stable pulsed combustion behavior. It also provided reliable quantitative response for trace sulfur detection. The dual-EPC design helped clarify the relationship between ignition and analytical combustion. The optimized conditions provide an experimental basis for self-developed PFPD systems.

Key words: pulsed flame photometric detector （PFPD）, delayed sulfur emission, quantitative repeatability, sulfur detection limit, linear range

CLC Number:

O658

CHEN Junbin, CAI Haoyuan, DENG Zhihan, JIN Yujie, SUN Jianhai. Development and performance testing of a high-sensitivity pulsed flame photometric detector[J]. Chinese Journal of Chromatography, 2026, 44(9): 1059-1073.

Figures/Tables 22

Fig. 1 Structural configuration of the PFPD and schematic illustrations of its flame-propagation process and optical detection patha. overall structure of the detector and division of functional modules； b. schematic illustration of the flame propagation path； c. schematic illustration of the optical detection path.PFPD： pulsed flame photometric detector； PMT： photomultiplier tube.

Fig. 2 Comparison of flame front morphology and ·OH response characteristics at different effective combustion chamber lengths （L）a. representative temperature field distributions of flame fronts at different effective lengths； b. transient ·OH mass fraction response curves at the midpoint of the chamber axis at different effective lengths.Combustion chamber inner diameter： 2 mm； combustion chamber H2 flow： 2.0 mL/min； combustion chamber air flow： 3.5 mL/min.

Fig. 3 PFPD gas flow diagramEPC： electronic pneumatic control. The gas flow diagram was prepared with the assistance of Shandong Lunan Ruihong Chemical Instrument Co.， Ltd.， based on the actual experimental system.

Fig. 4 PFPD operating principle

Fig. 5 Simulated temperature-field evolution during flame propagation and self-extinction in the combustion chamberEffective length： 12 mm. Other conditions are the same as in Fig. 2. The times shown on the panels denote the elapsed time after ignition.

Fig. 6 Transient responses of ·OH mass fraction and temperature at the monitoring pointConditions are the same as in Fig. 5. The monitoring point is located at the midpoint of the chamber axis.

Table 1 Independent ignition test conditions for the ignition chamber

Parameter	Value
Combustion chamber H₂ flow/（mL/min）	0
Combustion chamber air flow/（mL/min）	0
Combustion chamber N₂ flow/（mL/min）	1.5
Ignition chamber H₂ flow/（mL/min）	0
Ignition chamber air flow/（mL/min）	16.0
PMT voltage/V	630
Trigger level/mV	500
Ignition voltage/V	6.28
Detector temperature/℃	150

Fig. 7 Time waveform of combustion pulses under different ignition chamber H2 flow conditionsCombustion chamber H2 flow： 0 mL/min； combustion chamber air flow： 0 mL/min； ignition chamber air flow： 16.0 mL/min.

Table 2 Gas flow ratios and combustion frequency at a constant ignition chamber H2 flow （8.0 mL/min） with stepwise increases in ignition chamber air flow

Ignition chamber air flow/（mL/min）	Ignition chamber H₂/Air	Combustion frequency/Hz
13	0.62	0.64
14	0.57	2.08
15	0.53	2.56
16	0.50	3.03
17	0.47	3.23
18	0.44	3.57
19	0.42	3.85

Table 3 Gas flow ratios and combustion frequency at a constant combustion chamber air flow （3.5 mL/min） with stepwise increases in combustion chamber H2 flow

Combustion chamber H₂ /（mL/min）	Combustion chamber H₂/Air	Combustion frequency/Hz
0	0	4.00
1.0	0.29	4.17
2.0	0.57	4.17
3.0	0.86	4.35
4.0	1.14	4.35
5.0	1.43	4.76

Fig. 8 Time waveform of combustion pulses under different H2 flow conditions in the combustion chamberCombustion chamber air flow： 3.5 mL/min； ignition chamber H2 flow： 8.0 mL/min； ignition chamber air flow： 19.0 mL/min.

Fig. 9 Time waveform of sulfur luminescence response under oxygen-deficient conditionsCombustion chamber H2 flow： 5.0 mL/min； combustion chamber air flow： 3.5 mL/min； ignition chamber H2 flow： 8.0 mL/min； ignition chamber air flow： 19.0 mL/min.

Fig. 10 Three-dimensional overlay of sulfur luminescence response at different combustion chamber air flow ratesCombustion chamber H2 flow： 5.0 mL/min； ignition chamber H2 flow： 8.0 mL/min； ignition chamber air flow： 19.0 mL/min

Fig. 11 Heat map of peak area， detection sensitivity， and signal-to-noise ratio （SNR） as a function of combustion-chamber air flow rate （Data normalized）

Fig. 12 Three-dimensional overlay of sulfur luminescence response at different detector temperaturesCombustion chamber H2 flow： 5.0 mL/min； combustion chamber air flow： 3.7 mL/min； ignition chamber H2 flow： 8.0 mL/min； ignition chamber air flow： 19.0 mL/min.

Fig. 13 Heat map of peak area， detection sensitivity， and SNR as a function of temperature （data normalized）

Fig. 14 Three-dimensional overlay of sulfur luminescence response at different PMT voltagesConditions are the same as in Fig. 12.

Fig. 15 Peak height， SNR， and baseline noise versus PMT voltageThe error bars on the peak height and SNR bars represent the standard deviation （SD） of repeated measurements （n=7）. All error bars are shown on the original measurement scale without any amplification.

Fig. 16 Time-resolved delayed luminescence response curves of hydrocarbons and sulfur in the PFPD

Fig. 17 Linear calibration curve of sulfur-channel peak area versus H2S amount fraction for the PFPD

Fig. 18 Chromatograms of seven consecutive injections of 1.1 μmol/mol H₂S

Table 4 Comparison of the main performance metrics of the developed PFPD and commercial sulfur detectors

Detector	Sulfur detection limit/（g/s）	Linear range	Quantitative repeatability
PFPD developed in this work	4.1×10^-13	2 orders of magnitude	0.37% （1.1 μmol/mol H₂S）
5383 PFPD^{［17，19］}	1×10^-12	2.4 orders of magnitude	11% （0.657 μmol/mol H₂S）
Nexis FPD-2030^{［20，21］}	2×10^-12	3 orders of magnitude	0.91% （1 μmol/mol thiophene）

References

[1]	Cao J， Sun Q， Zhao D， et al. J Hazard Mater， 2020， 385： 121511
[2]	Ou Y， Wu M， Yu Y， et al. Sci Total Environ， 2024， 929： 172478
[3]	Spatolisano E， Restelli F， Pellegrini L A， et al. Energies， 2024， 17（3）： 620
[4]	Zhang X， Qiu X， Wang R. Catalysts， 2024， 14（12）： 952
[5]	Ma L， Zhao R， Li J， et al. Environ Pollut， 2024， 350： 123946
[6]	Chen Z X， Zhou L， Wu S H， et al. Natural Gas Industry， 2025， 45（5）： 150
	陈梓浔，周理，吴姝虹，等. 天然气工业， 2025， 45（5）： 150
[7]	Syty A， Dean J A. Appl Opt， 1968， 7（7）： 1331
[8]	Fair R W， Thrush B A. Trans Faraday Soc， 1969， 65： 1208
[9]	Cheskis S， Atar E， Amirav A. Anal Chem， 1993， 65（5）： 539
[10]	Amirav A， Jing H. Anal Chem， 1995， 67（18）： 3305
[11]	Jing H， Amirav A. J Chromatogr A， 1998， 805（1/2）： 177
[12]	Kalontarov L， Jing H， Amirav A， et al. J Chromatogr A， 1995， 696（2）： 245
[13]	Aue W A， Eisener C G， Gebhardt J A， et al. J Chromatogr A， 1995， 699（1/2）： 195
[14]	Catalan L J J， Liang V， Jia C Q. J Chromatogr A， 2006， 1136（1）： 89
[15]	Massimi N， Polvara E， Invernizzi M， et al. Chem Eng Trans， 2024， 112： 91
[16]	Torres-Herrera S， González-Cortés J J， Almenglo F， et al. J Hazard Mater， 2022， 424： 127667
[17]	Analytical OI. 5383 PFPD Pulsed Flame Photometric Detector Operator Manual. ［2026-02-28］. https://www.ysi.com/File%20Library/Documents/Manuals/Pulsed-Flame-Photometric-Detector-Operator-Manual.pdf
[18]	GB/T 32190-2015
[19]	Elmore C， Duffy M. Am Lab， 2016， 48（2）： 20
[20]	Corporation Shimadzu. Nexis GC-2030 Specifications. ［2026-04-12］.https://www.shimadzu.com.sg/an/sites/shimadzu.com.sg.an/files/pim/pim_document_file/sap_an/spec_sheet/25256/Specs_Nexis_GC-2030_%5Bc184e042E%5D.pdf
[21]	Corporation Shimadzu. Analysis of Thiophene in Benzene. ［2026-04-12］. https://www.shimadzu.com/an/sites/shimadzu.com.an/files/pim/pim_document_file/applications/applica tion_note/13392/jpc117007.pdf