Refinement and Application of the Absorption Cross-Section of Molecular Bromine with Relevance to Atmospheric Chemistry

Abstract

Bromine (Br₂) plays a crucial role in atmospheric oxidative chemistry, particularly in polar, marine, and stratospheric environments, where it significantly impacts ozone depletion and mercury oxidation. In this study, high-resolution measurements of the Br₂ absorption cross section were evaluated across 325–850 nm (11765 to 30,769 cm^–1), with emphasis on features due to the B³Π_ou ← X¹Σ_g ⁺ transition at 530 nm (18,867 cm^–1). UV–vis spectra revealed well-resolved vibronic features, enabling the accurate assignment of absorption peaks. Simulated rovibronic spectra reproduced the measured features and confirmed spectral assignments. The data set produced in this work shows improved signal-to-noise and peak distinction compared to prior studies and reveals differences of up to 25% relative to literature values, and to 6.7% relative to the JPL evaluation values, in structured regions of the spectrum, which directly alters calculated Br₂ photolysis rates used in atmospheric chemistry models. These refined cross sections serve as robust reference data for laboratory and atmospheric research, supporting improved photochemical modeling of reactive halogen species. A broadband cavity-enhanced absorption spectroscopy (BBCEAS) instrument was employed to test the application of the evaluated absorption cross-section in a field-ready instrument and setting. The BBCEAS method achieved a detection limit of 97 ppt and an effective optical path length of 22.6 km, facilitating sensitive detection suitable for field applications.

1. Introduction

Bromine (Br₂) chemistry plays a critical role in atmospheric processes, particularly in ozone depletion and the oxidative balance of the atmosphere. This is especially evident during polar spring and in the midlatitude stratosphere. − Although bromine is less abundant than other halogens such as chlorine and fluorine, its high oxidative potential gives it a disproportionately significant impact on atmospheric chemistry. , Br₂ is approximately 60–65 times more effective than chlorine at destroying ozone molecules per atom within catalytic cycles, making it a key driver of ozone depletion in both the stratosphere and troposphere. −

In the stratosphere, bromine released from anthropogenic compounds such as halons and methyl bromide participates in catalytic cycles that deplete ozone. − This ozone loss contributes to increased ultraviolet (UV) radiation at Earth’s surface. Despite its lower atmospheric abundance compared to chlorine, bromine’s high reactivity enables it to account for a notable fraction of ozone loss. In the troposphere, Br₂ chemistry is particularly important in polar regions, where “bromine explosion” events are observed during the polar spring. , These events, triggered by the release of bromine from sea salt aerosols and frost flowers on sea ice, result in localized but rapid ozone destruction in the lower atmosphere. Comparable halogen-driven processes have also been reported for saline inland lakes, including the Great Salt Lake in Utah. ,

Such phenomena underscore the complexity of bromine chemistry, with implications for ozone levels and trace gas lifetimes. Beyond its role in ozone chemistry, bromine species significantly impact the oxidation of mercury and other trace gases. , Bromine radicals (Br and BrO) oxidize elemental mercury (Hg⁰) into more reactive and soluble forms that are subsequently deposited in ecosystems, contributing to environmental mercury contamination. These processes, known as mercury depletion events, is particularly pronounced in polar regions. Furthermore, bromine affects the oxidative capacity of the atmosphere by altering the balance of key radicals, such as hydroxyl (OH) and hydroperoxyl (HO₂), which drive the removal of pollutants and greenhouse gases. Interactions with VOCs and halogenated species further extend its influence in the atmosphere. Understanding Br₂ chemistry is essential to elucidate its role in ozone depletion, mercury oxidation, and atmospheric chemical cycles. This understanding is key to evaluating its impact on air quality, climate and ecosystems. To assess the atmospheric impact of bromine chemistry, it is necessary to quantify the abundance of its molecular precursors. Br₂ serves as an important reservoir and photolabile source of reactive bromine radicals, and therefore determining its atmospheric concentration is essential for evaluating its role in ozone loss and oxidative processes. Optical absorption measurements provide a direct method for determining Br₂ concentrations through the Beer–Lambert law. However, the accuracy of such measurements depends critically on the quality of the absorption cross-section used in the retrieval. Consequently, improved and well-characterized Br₂ absorption coefficients are required in order to reliably convert measured optical extinction into atmospheric abundance. Because Br₂ photolysis frequencies are directly proportional to the absorption cross-section, errors in σ­(λ) also propagate into calculated radical production rates used in atmospheric models.

In this study, we evaluate the absorption cross-section of Br₂ from 325–850 nm (11,765 to 30,769 cm^–1). Although prior measurements exist, this study provides higher-resolution, signal-to-noise, and continuous coverage across a broader wavelength range of the Br₂ absorption cross-section than any previously published data, leveraging an instrument with a lower detection limit to produce a low-noise, high-resolution spectrum. We compare our results with previously obtained high-resolution Br₂ absorption cross sections that were made available through personal communication. Additionally, we evaluate the use of the structure in the absorption cross-section from the B³Π_ou ← X¹Σ_g ⁺ transition at 530 nm (18867.9 cm^–1) for field measurements using broadband cavity-enhanced absorption spectroscopy (BBCEAS).

2. Methods

2.1. Ultraviolet–Visible (UV/Vis) Spectrophotometer Experimental Setup and Measurement

An Ultraviolet–visible (UV/vis) spectrophotometer (Jasco V-770) was used to measure Br₂ absorption in the 325–850 nm (11,765 to 30,769 cm^–1) range, with data collected at 0.025 nm intervals and a 0.1 nm bandwidth. The spectrophotometer has a wavelength accuracy of ±0.3 nm, wavelength reproducibility of ±0.05 nm, and photometric accuracy of ±0.0015 absorbance units. The Br₂ sample was placed in a glass cell with silica windows, which was first evacuated and then backfilled with 20 Torr (2.67 kPa) of Br₂ (approximately 10% of the vapor pressure of Br₂). The optical path length of the cell was 10 cm. This setup provided a high-resolution spectrum over a broad wavelength range, allowing for the assignment of the calculated line spectrum to specific peaks. To obtain a spectrum where peaks due to transitions could be identified and differentiated from noise, 100 scans of Br₂ taken in the spectrophotometer were baseline corrected and averaged. Repeated scans showed good reproducibility, and the uncertainty in the measured Br₂ absorption cross-section was estimated from contributions due to pressure measurement, temperature, detector noise, and baseline correction.

2.2. Calculation of Br₂ Rovibrational Line Spectrum and Simulated Absorption Cross-Section

High-resolution rovibronic transition frequencies for Br₂ were calculated for the B³Π_ou ← X¹Σ_g ⁺ electronic transition across all bound vibrational levels. Calculations were performed for the three naturally occurring isotopologues, ⁷⁹Br₂ (25.7%), ^79,81Br₂ (49.9%), and ⁸¹Br₂ (24.3%), using isotopologue-specific vibrational and rotational constants.

The rovibrational energy levels of the X¹Σ_g ⁺ state were calculated using Dunham expansion

$$E^{″}(v^{″},J^{″})=\sum _{k,m}{Y}_{k,m}{(v^{″}+\frac{1}{2})}^k{[J^{″}(J^{″}+1)]}^m$$ 1

where v″ is the vibrational quantum number and J″ is the rotational quantum number of the X state. Dunham constants Y _m,k were taken from Fosca et al., including terms over the following ranges: m = 0 (k = 1–6), m = 1 (k = 0–4), and m = 2 (k = 0–2) for all isotopologues. Calculations were carried out for vibrational states up to v″ = 2 and rotational states up to J″ = 52, just prior to dissociation.

Energies of the B³Π_ou rovibrational states were calculated using the same Dunham-type expression with m = 0–1 and k = 0–13. Constants for the excited state were taken from Gerstenkorn and Luc. They did not report higher order Dunham constants; however, they reported higher order rotational corrections which we included in our excited energy state calculation by adding

$${E_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{r}}(v^{\prime },J^{\prime })=D}_{v\prime }{[J^{\prime }(J^{\prime }+1)]}^{\prime 2}+H_{v\prime }{[J^{\prime }(J^{\prime }+1)]}^3+L_{v\prime }{[J^{\prime }(J^{\prime }+1)]}^4+M_{v\prime }{[J^{\prime }(J^{\prime }+1)]}^5$$ 2

These constants are specific to each vibrational level of the B state and were reported by Fosca et al., Gerstenkorn and Luc, and Gerstenkorn et al. The total rovibrational energy of the excited state was therefore

$$E^{\prime (v^{\prime },J^{\prime })}=T_{\mathrm{ee}}+\sum _{k=\mathrm{0-13}}\sum _{m=\mathrm{0-1}}{Y}_{k,m}{(v^{\prime }+\frac{1}{2})}^k{[J^{\prime (J^{\prime }+1)}]}^m+E_{\mathrm{corr}}(v^{\prime },J^{\prime })$$ 3

where T _ee is the electronic term origin for the B³Π_0u state taken from the National Institute of Standards and Technology Chemistry WebBook. The reported value of T _ee was 15,763.2 cm^–1 and was combined with the Dunham expansion and higher-order rotational corrections to calculate the total excited-state rovibrational energies. Uncertainty in T _ee is small relative to the overall uncertainty in the calculated transition positions and is estimated to be less than 1 cm^–1.

Transition energies for the P- and R-branches were obtained from differences between excited and ground states

$${\nu }_{\mathrm{P}}=E^{\prime }(v^{\prime },J^{\prime }-1)-E^{″}(v^{″},J^{\prime })$$ 4

$${\nu }_{\mathrm{R}}=E^{\prime }(v^{\prime },J^{\prime }+1)-E^{″}(v^{″},J^{\prime })$$ 5

where v′, J′ and v″, J″ denote the upper and lower vibrational and rotational quantum numbers, respectively.

Relative line intensities were calculated from ground state rotational populations at T = 298 K

$$N(J)=(2J+1){\mathrm{e}}^{-E_{\mathrm{r}\mathrm{o}\mathrm{t}}/\mathit{kT}}$$ 6

where k is the Boltzmann constant and E _rot (J) is the rotational energy of the ground electronic state,

The rotational energy used in the Boltzmann factor was

$$E_{\mathrm{rot}}=E_{v,J}^{″}-E_{v,J=0}^{″}$$ 7

Nuclear spin statistical weights (N _ns) were included for the homonuclear isotopologues (⁷⁹Br₂, ⁸¹Br₂) where bromine nuclear spin angular moment is I = 3/2. For the homonuclear species N _ns = I(2I + 1) = 6 for even J states and N _ns = (I + 1)­(2I + 1) = 10 for odd J states.

The probability of a given rovibrational state was found by the ratio of these two terms and the partition function according to

$$P(\mathrm{J})=\frac{N(\mathrm{J})\times N_{\mathrm{n}\mathrm{s}}}{Q_{\mathrm{r}\mathrm{o}\mathrm{t}}{Q}_{\mathrm{n}\mathrm{s}}}$$ 8

Where Q _ns = (2I + 1)² and $Q_{\mathrm{rot}}=\frac{\mathit{kT}}{\sigma B_v}$ where σ the symmetry number is 1 for the heteronuclear isotopologue and 2 for homonuclear isotopologues. For the heteronuclear isotopologue (^79,81Br₂), the nuclear spin terms are not included.

Equation determines the relative intensity of the various rovibronic transitions belonging to a given band; however, to compare the intensity of different bands to each other requires the addition of the population of the ground electronic vibrational states as well as the Franck–Condon factors for the vibronic transitions. The vibrational state population was determined from the Boltzmann distribution according to

$$\omega ({\upsilon }^{″})=\exp [-\frac{E^{″}({\upsilon }^{″})\mathit{hc}}{k_{\mathrm{B}}T}]$$ 9

where h is Planck’s constant, c is the speed of light, k _B is Boltzmann’s constant, and T is the temperature and E″(υ^″) is the vibrational energy of the υ^″ state.

The normalized fractional population of each vibrational level was

$$P({\upsilon }^{″})=\frac{\omega ({\upsilon }^{″})}{\sum _u\omega (u)}$$ 10

Franck–Condon factors from Coxon and Barrow were used to weight vibrational transitions. While we acknowledge that the Franck–Condon Factors decrease with separation as the state increases, for v′ > 44, the Franck–Condon factor for v′ = 44 was applied as an approximation as there is no experimental or observed data available above v′ = 44 in addition to these transitions being very small due to the small Boltzmann populations in these states. The final intensity of each line combined isotopic abundance (25.69% (⁷⁹Br₂), 24.31% (⁸¹Br₂) and 49.99% (^79,81Br₂)), rotational weighting, vibrational population, and Franck–Condon factors. Stick spectra for each isotopologue were generated in MATLAB and then isotopically weighted and binned for clarity when compared to measured absorption cross sections. The calculated transition energies are expected to be accurate to better than 1 cm^–1 based on the reported uncertainties and fit residuals associated with the Dunham constants and excited-state corrections used from the literature. Relative intensities are expected to be more uncertain due to approximations in Franck–Condon factors, partition functions, and the treatment of higher vibrational levels above ν′ = 44, but remain sufficient for assignment of the observed rovibronic structure. A complete list of calculated rovibrational transitions and band assignments is provided on BYU Scholar’s Archive.

2.3. BBCEAS Experimental Setup and Measurement

The optical cavity used in this study consisted of two 25 mm diameter highly reflective dichroic mirrors (>99.99% reflectivity) optimized for 520 nm (19,230 cm^–1), with a radius of curvature of 1000 mm (Layertec, 140,964). These mirrors were mounted in a 3D-printed cage assembly and positioned 96 cm apart. The cage assembly consisted of carbon-fiber rods held in place by 3D-printed brackets that provided stable mirror alignment while minimizing weight and cost. The 3D-printed parts, fabricated from polylactic acid (PLA) on a desktop printer, served only as structural supports and did not contact the optical path. Stainless steel inserts and O-rings were used at the mirror mounts to ensure proper sealing and alignment. This type of modular cage system has been employed in similar BBCEAS instruments to provide mechanical stability, reproducibility of alignment, and flexibility for rapid modifications during instrument development. −

The light source for the BBCEAS instrument was a high-powered LED (Mouser Electronics, 720-LZ1–10G1020000G2) with a 0.37 W light output centered at 523 nm (19,120 cm^–1). The LED was mounted on a copper heat sink and temperature-controlled using a Peltier cooler (Digikey, CP30138) and a temperature controller (Meerstetter Engineering TEC-1091). Light emitted by the LED was directed into the cavity using collimation optics and, on the far side of the cavity, collected and focused on to a fiber optic (ThorLabs, BFL200HSO2) using focusing optics.

To isolate the wavelengths of interest and minimize stray light, a 550 nm short-pass filter (ThorLabs, FESH0550), with an OD of >5 above 543 nm, and a 2 mm thick OG515 Schott yellow glass filter (Thorlabs, FGL515) were employed. The fiber optic directed the light to a spectrograph for detection (Andor SR-303i Spectrograph with a DU440-BV CCD camera). The CCD was cooled to −20 °C to reduce noise, and a 1200 grooves/mm grating combined with a 25 μm entrance slit provided a full-width at half-maximum (FWHM) resolution of 0.22 nm.

Spectra were acquired every minute by averaging 60 1-s integrations. Further instrument details are described elsewhere. − A schematic of the complete experimental setup is shown in Figure .

1.

Experimental setup showing BBCEAS.

When compared to the UV–vis spectrophotometer, the BBCEAS instrument offers faster response and adequate resolution for field deployment in a compact and easy-to-transport setup, motivating the use of both instruments in this study. The UV/vis spectrophotometer was used to define the Br₂ absorption cross-section across the full wavelength range, while the BBCEAS instrument was used to evaluate how the structured absorption features near 530 nm could be applied in a field-deployable measurement approach.

Bromine vapor was introduced into the BBCEAS optical cavity by flowing ultrahigh purity nitrogen gas (AirGas, > 0.99998) through the headspace of a flask containing liquid bromine. Flow rates of 0, 50, 75, and 100 standard cubic centimeters per minute (SCCM) were used. The bromine-containing flow was then mixed with a pure nitrogen flow for dilution, ensuring a constant total flow of 3000 SCCM into the instrument. Flow rates were precisely controlled using a mass flow controller (MKS Multi Gas Controller 647C).

The flask containing liquid bromine was immersed in a temperature-controlled slurry of acetone and dry ice to maintain stable conditions at −78 °C. Temperature control was critical to prevent bromine from condensing in the system. To ensure this, the vapor pressure and saturation concentration of bromine were calculated using the Antoine Equation with coefficients provided by the National Institute of Standards and Technology (NIST). These calculations verified that condensation would not occur under experimental conditions and are shown in Figure and Table S3.

2.

Calculated saturation concentration of bromine for a given temperature and calculated vapor pressure for bromine for a given temperature.

3. Results and Discussion

3.1. Defining the Br₂ Absorption Cross-Section Using UV/Vis Spectrophotometer

Table presents a comparison of the different available absorption cross sections both from the literature and from databases. The JPL Evaluation is primarily based on a Gaussian parametrization of the absorption cross-section data reported by Maric et al. at 5 nm intervals and therefore does not capture the fine rovibronic structure resolved in the present work. The data set presented in this work yields the widest measured window (525 nm), with low noise, and with high enough resolution to see the structure in the measured absorption cross-section. The discussion about the data set presented in this work will focus on the comparison to the absorption cross-section identified by the JPL evaluation. Absorption cross sections are direct inputs to photolysis calculations. Because photolysis frequencies are proportional to σ­(λ), uncertainties in the Br₂ cross-section translate directly into uncertainties in calculated Br₂ photolysis rates.

1. Comparison of Room Temperature (293-303 K) Available Br₂ Absorption Cross Sections .

authorship	wavelength range (nm)	average point spacing in 500–600 nm region (nm)	fine structure in 500–600 nm region	type of publication	publication year	% difference from this work (500–572 nm)
Acton et al.	327–571	14	No	research article	1936	24.8
Gray et al.	240–579	>33	No	research article	1930	too few points to integrate
Hemenway et al.	570–616	N/A	No	research article	1979	too few points to integrate
Hubinger et al.	190–600	10	No	research article	1995	13.3
IUPAC	200–500	5	No	research article	2007	outside wavelength range
JPL	200–650	5	No	JPL Evaluation	2011	6.7
Maric et al.	200–650	1	No	research article	1994	5.9
McMillan et al.	200–600	1	Yes	personal communication	1966	14.0
Passchier et al.	200–230	N/A	No	research article	1967	outside wavelength range
Passchier et al.	300–750	10	No	research article	1967	–0.5
Roxlo et al.	170–230	N/A	No	research article	1980	outside wavelength range
Schneider et al.	315–512	0.2	Yes	personal communication	1991	outside wavelength range
Schneider et al.	504–553	0.01	Yes	personal communication	1991	Personal communication only
Seery et al.	320–590	10	No	research article	1964	17.5
Tellinghuisen et al.	190–300	N/A	No	research article	2008	outside wavelength range
Wen et al.	220–290	5	No	research article	1972	outside wavelength range
Zaraga et al.	558	N/A	No	research article	1976	too few points to integrate
Flowerday et al.	325–850	0.025	Yes	this work	2025	outside wavelength range

^aThis table includes a comparison of integrated absorption cross sections defined in this work to those defined in literature. The structured region of the absorption cross-section that included the greatest number of studies was chosen for this comparison.

Figure shows the absorption cross-section for Br₂ defined using the UV–vis in this work overlaid on the Br₂ absorption cross-section defined in the JPL Evaluation. The spectra are in good agreement; however, the JPL Evaluation lacks the resolution needed to resolve fine structure between 510 and 610 nm as seen in Figure B. This is because the absorption cross-section presented by the JPL evaluation is recorded at a resolution of 5 nm and this work recorded data at a resolution of 0.025 nm. The observed differences are strongly influenced by spectral resolution. Lower-resolution data sets smooth over the fine rovibronic structure present in the 500–572 nm region, reducing the apparent peak intensities and altering the integrated absorption cross-section. In contrast, the 0.025 nm spacing used in this work preserves the detailed band structure, allowing more accurate representation of narrow absorption features that are averaged out in lower-resolution data sets such as the JPL Evaluation. The uncertainty in the measurement of Br₂ in this work was calculated in quadrature as a combination of the error in the temperature measurements (0.1%), baratron measurement (1%), detector noise (wavelength dependent) and found to be an average of 2% in the structured region of the spectrum (500–572 nm, 17,482–20,000 cm^–1) and 1.0% total error from 400 to 500 nm (20,000–25,000 cm^–1). See Figure S1 for noise determination and error propagation.

3.

(A) Br₂ absorption cross sections defined in this work, in red, overlaid on the JPL Evaluation, in green. (B) Zoomed in structured region of the Br₂ absorption cross sections defined in this work, in red, overlaid on the JPL Evaluation, in green.

To directly compare the absorption cross-section defined in this work with the absorption cross sections defined in literature and to the personal communications found in the MPI Mainz Spectral Database, absorption cross sections were integrated from 500 to 572 nm. This range was chosen to maximize overlap with available studies while including structured absorption features. The integration was done using Simpson’s Rule, the trapezoidal rule, a left Riemann sum, and a cumulative sum for comparison. Results from Simpson’s Rule are presented in Table . Additional integration results obtained using the trapezoidal and Riemann methods are summarized in Tables S1–S2 of the Supporting Information. The highest percentage difference is in comparison to Acton et al. where there was a 24.8% difference, and the smallest percentage difference is in comparison to Passchier et al. at 0.5%. The percentage difference from the JPL evaluation is 6.7%, which could prove to be a significant difference when implemented in atmospheric models and should be evaluated by the modeling community. Because Br₂ photolysis frequencies scale directly with the absorption cross-section, a ∼6–7% difference from the JPL Evaluation would produce a comparable change in calculated Br₂ photolysis rates under identical radiation fields.

The uncertainty in the absorption cross-section measured in this work was estimated to be approximately 2% in the structured region and approximately 1% in the higher-absorption region. These uncertainties are substantially smaller than the observed 6.7% integrated difference relative to the JPL Evaluation, suggesting that the difference is not solely attributable to measurement error and likely reflects the improved spectral resolution and greater ability to resolve fine structure in the present data set. Because Br₂ photolysis frequencies scale linearly with the absorption cross-section, a 6–7% increase in σ­(λ) would directly increase calculated Br₂ photolysis rates and reactive bromine production by a similar amount under identical radiation fields.

3.2. Calculation of Br₂ Rovibrational Transition Peaks

Rovibrational spectra for the three isotopologues (⁷⁹Br₂, ^79,81Br₂, ⁸¹Br₂) were calculated over 15,000–20,000 cm^–1. The calculated relative populations for the υ^″ vibrational state showed no significant population past υ^″ = 2 at the temperature of this measurement. These relative populations can be found in Table .

2. Calculated Boltzmann Distribution of the Starting States of Molecules in Percentage of the Population.

υ″	% population
0	78.73
1	16.70
2	3.58
3	0.78
4	0.17
5	0.04

The calculated B–X rovibronic lines were compared with the measured absorption cross-section spectrum to validate the structured B–X component of the spectrum. Figure (A) shows the region of the absorption cross-section spectrum where structure from the B-X transition exists and the experimental result is overlaid with the calculated lines. The P- and R-branch peaks produce a band head, resulting in the peak shapes observed in the measured spectrum. As seen in calculated ultrahigh resolution stick spectrum, R-branch peaks pile up and turn around forming a band head. This band head can also be seen in the UV–vis spectrum where the band has a sharp edge on the blue side of each band and is red degraded, tailing off at lower frequencies. This is the result of Bromine having a smaller rotational constant in the excited state than it does in the ground state. This behavior was consistent across the structured region, confirming that the observed spectral features are real and arise from specific rovibronic transitions. The esolved rovibronic band-head structure in the measured absorption cross-section extend beyond the band heads, but this can be explained by the added intensity of overlapping rovibronic lines in the denser regions, where one band begins as the previous one dies out. Figure (B) shows the same comparison as Figure (A), with vibronic labels indicating the ranges of the vibronic band. Using an example from Figure (B) to show this phenomenon, the 2 → 16 band fades as the 1 → 16 band grows in, and the combined intensity of their overlap makes the measured absorption peak appear shifted relative to the calculated band head. Extended simulated band structures and full transition tables are available in the Supporting Information (Figure S2 and Supporting data sets).

4.

Comparison of calculated and measured Br₂ rovibrational structure in the 500–625 nm (16,000–20,000 cm-1) region. This Figure is presented in wavenumber units because the rovibrational transition calculations, Dunham constants, and spectroscopic assignments are conventionally expressed in cm^–1. (A) The structured region of the defined absorption cross-section from this work (orange) is overlaid with the line spectrum of the calculated rovibrational transitions for ^79,81Br₂ (red), ⁷⁹Br₂ (blue), and ⁸¹Br₂ (black). (B) A zoomed-in portion of panel A showing individual rotational progressions in a vibronic transition are labeled in the form of ν″ → ν′. (C) Gaussian-convolved B–X line spectrum overlaid with the measured spectrum; differences in baseline/envelope reflect an additional smooth continuum contribution (C–X) not included in the line-list model.

Rovibrational stick spectra for Br₂ were calculated and binned onto a uniform 0.025 nm wavelength grid to match the sampling of the measured cross-section data. Instrumental broadening was applied by convolving the binned stick spectrum with a Gaussian line-shape function having a FWHM of 0.1 nm, consistent with the experimental spectral bandwidth. The convolved model spectrum was then directly compared to the measured Br₂ absorption cross section on a common wavelength axis without additional scaling to assess agreement in band structure and envelope shape. The convolved line spectrum is plotted with the measured spectrum and shown in Figure (C).

Prior analyses of the Br₂ visible absorption continuum show that absorption is dominated by a smooth continuum contribution associated with excitation into a repulsive ¹Π_u state (sometimes denoted C–X), with additional absorption from B–X contributions at lower energies; importantly, the C–X contribution is continuum-like and does not generate discrete rovibronic band-head structure. , However, absorption due to the C-X state does not start until 20,000 cm^–1 and, therefore, does not contribute to absorption in this structured region of interest.

Figure (C) shows the comparison between the measured data and the convolution of the calculated line spectrum. This comparison shows good agreement between the measured spectrum and the convolved calculated spectrum. Residual disagreement in the absolute baseline is potentially due to the use of relative intensities rather than absolute intensities used in the calculations as well as to the approximation of FCF > 44 as being the same as that of 44 given that there is no experimental data available above v′ = 44.

3.3. Defining the Br₂ Absorption Cross-Section Using BBCEAS

While the UV/vis spectrophotometer measurements were used to define the Br₂ absorption cross-section presented in this work, the BBCEAS measurements were used to demonstrate the application of the structured 530 nm region for sensitive field-based Br₂ detection.

The mirrors forming the optical cavity exhibited a peak reflectivity of 0.99997, corresponding to an effective optical path length of 22.6 km at 548 nm (18,248 cm^–1). This extended path length enabled a 1 min detection limit of 97 ppt, based on the root-mean-square (RMS) of the fit residual shown in Figure (A). The RMS residual shown in Figure (A) is comparable to those reported in previous BBCEAS studies for trace gas retrievals under laboratory conditions and corresponds to approximately 2% of the peak extinction signal. This indicates that the fitting procedure reproduces the measured extinction spectrum with high fidelity and is consistent with fit quality reported for previous BBCEAS measurements of trace gases. − This sensitivity facilitated the measurement of Br₂ concentrations during the experiment, as illustrated in Figure (B) and is ideal for field measurements. The concentrations shown in Figure (B) were derived by fitting the measured BBCEAS extinction spectra using the Br₂ absorption cross-section evaluated in this work over the 515–550 nm region.

5.

(A) Measured BBCEAS extinction spectrum (red) and fit to data (black), with fit residual shown above, (B) Time series of Br₂ concentration derived by fitting the measured BBCEAS using the evaluated Br₂ absorption cross-section over the 515–550 nm region.

The uncertainty in the measurements by the BBCEAS is greater at wavelengths below 515 nm, demonstrated in noise on the fit residual in Figure (A), as the Schott glass filter only allows 50% of the light through below 515 nm, however, the LED is centered at 523 nm and there is sufficient light to make a measurement below 515 nm. The experimental error was determined by combining the noise on the spectrum with the noise on the detector in quadrature, yielding an uncertainty of up to 3%, depending on wavelength (2% on average), for the Br₂ measurements made by the BBCEAS. Uncertainty values were validated through repeated trials under identical flow and temperature conditions.

4. Conclusions

In this study, we present a measurement of the Br₂ absorption cross-section from 325 to 850 nm (11765 to 30769 cm^–1) using a UV/vis spectrophotometer and focused on the B³Π_ou ← X¹Σ_g ⁺ electronic transition at 530 nm (18867.9 cm^–1) as a region ideal for field application using broadband cavity-enhanced absorption spectroscopy (BBCEAS). The absorption cross sections described in this work show excellent agreement with previously reported spectra and offers significantly improved noise characteristics and spectral resolution. The observed structure was further confirmed by comparison with calculated rovibronic spectra, ensuring that the measured features arise from true molecular transitions. Comparisons with literature data through integrated cross-section values highlight the changes that spectral resolution can make when interpreting photolysis values in these regions, highlight the changes that spectral resolution can make when interpreting photolysis values in these regions, with differences of up to 25% relative to literature values and 6.7% relative to the JPL Evaluation values. Because photolysis frequencies scale linearly with the absorption cross-section, the 6–7% difference relative to the JPL Evaluation identified here would directly alter modeled Br₂ photolysis rates and the production of reactive bromine radicals, thereby affecting simulated ozone loss efficiency and atmospheric oxidative capacity in both regional and global models. Although a 6–7% change may appear modest, such a change propagates directly into Br₂ photolysis frequencies and therefore into modeled Br atom production, BrO formation, and catalytic ozone loss efficiency. Over long model integrations or in bromine-sensitive environments such as polar regions, the marine boundary layer, or saline lake systems, such differences may influence predicted oxidant budgets and ozone depletion rates. Increased Br atom production enhances BrO formation and catalytic halogen ozone loss cycles, thereby modifying modeled oxidative capacity and ozone depletion efficiency. These refined measurements provide a reference spectrum that supports atmospheric modeling and field-based detection of reactive bromine species, especially in regions where Br₂ significantly drives ozone depletion and trace gas oxidation. Field application results demonstrate that BBCEAS provides a sensitive and reliable method for characterizing Br₂ absorption features, with a detection limit of 97 ppt and an effective optical path length exceeding 22 km. In parallel, the spectrophotometer provided higher-resolution structural information that was not accessible by BBCEAS, making the combination of the two methods essential for both detailed laboratory evaluation and characterization and field deployment.

Available on BYU’s ScholarsArchive https://scholarsarchive.byu.edu/data/84 (accessed on 6 June 2025).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsenvironau.6c00050.

• Tables of integrated Br₂ absorption cross sections and percentage differences using multiple numerical integration methods; table of calculated Br₂ vapor pressures and saturation concentrations; figure showing noise determination and error characterization; figure showing extended calculated rovibrational transition overlays; link to large supplemental transition and cross-section data sets (PDF)

C.E.F: Conceptualization, writing of the original manuscript, data acquisition, data analysis. R.T.: Conceptualization, data analysis, writingreview and editing, funding acquisition. E.T.S: Data analysis, writing review and editing. M.C.A: Data analysis, writingreview and editing. J.J.S.: data acquisition, reviewing and editing. J.C.H.: Conceptualization, writingreview and editing, supervision, and funding acquisition. All Authors have read and agreed to the published version of the manuscript.

This work was supported by the National Science Foundation, grant #2321381.

The authors declare no competing financial interest.