Database of low‐temperature absorption and fluorescence spectra of native photosynthetic tetrapyrrole macrocycles

Abstract

Low‐temperature (77 K) absorption and fluorescence spectra of 12 naturally occurring photosynthetic tetrapyrrole macrocycles have been recorded in a frozen glass (2‐methyltetrahydrofuran). The compounds encompass distinct chromophore classes: porphyrin, chlorophyll c ₂; chlorin, chlorophylls a, b, d, f and bacteriochlorophylls c, d, e, f; and bacteriochlorin, bacteriochlorophylls a, b, g. The spectra are compared with those of the same pigment in liquid solution (predominantly 2‐methyltetrahydrofuran) at room temperature (293 K). The measured Stokes shifts at 77 K across the 12 macrocycles range from ~30 to 300 cm⁻¹. The spectral data in digital form are made available as part of the PhotochemCAD databases. Literature searches have revealed extensive published data for Chl a (often in biological matrices) but at best rather limited data for less common macrocycles. The availability of a systematic collection of curated spectral data collected at low temperature should be useful for a variety of assessments, including reconstruction of absorption spectra of (bacterio)chlorophyll‐containing protein complexes, vibrational analysis of absorption and fluorescence spectra, and calculations where knowledge of energy levels is important.

The absorption and fluorescence spectra at 77 K (frozen glass) and room temperature (fluid solution) are provided for 12 native photosynthetic macrocycles encompassing one porphyrin, eight chlorins, and 3 bacteriochlorins.

Abbreviations

(B)Chls

all macrocycles described herein

2‐MTHF

2‐methyltetrahydrofuran

BChl

bacteriochlorophyll

Chl

chlorophyll

DE

diethyl ether

EPA

diethyl ether/isopentane/ethanol (5/5/2)

F

formyl

FMO

Fenna‐Matthews‐Olson

FRET

Förster resonance energy transfer

FWHM

full‐width‐at‐half‐maximum

HPLC

high performance liquid chromatography

LH

light‐harvesting

MeOH

methanol

V

vinyl

INTRODUCTION

Chlorophylls and bacteriochlorophylls are the most important and abundant photosynthetic pigments on Earth. The compounds have several important functions in photosynthesis, including the absorption of light, transfer of excitation energy in the antenna complexes, and separation of charge in reaction centers. ¹ , ² There are approximately one hundred chlorophyll structures known today, of which 12 are broadly utilized by various photosynthetic organisms. The degree of unsaturation of the macrocycle, which strongly influences the characteristic features of the absorption spectrum, provides a logical basis for division into three classes: (I) fully unsaturated macrocycle, porphyrin (Chl c ₂); (II) dihydroporphyrin (chlorin) macrocycle (Chls a, b, d, f and BChls c, d, e, f); and (III) tetrahydroporphyrin (bacteriochlorin) macrocycle (BChls a, b, g). The compounds are shown in Chart 1 and are collectively referred to as “(B)Chls.”

CHART 1.

Structures of native tetrapyrrole macrocycles studied herein.

Due to their ecological importance and widespread scientific interest, the absorption and fluorescence spectra of members of the photosynthetic tetrapyrroles have been recorded in numerous solvents and reported in countless publications. The spectra have been recorded chiefly at room temperature, although some studies have been carried out at low temperatures. Many studies have concerned pigment–protein complexes (e.g., reaction centers and antenna complexes); some include intact chromatophores, while others are for molecules in homogeneous solution. The richness of sample types and the very far‐flung nature of the scientific literature together have thwarted the assembly of systematic collections of spectral data. A recent database has been compiled of the absorption and fluorescence spectra of 150 chlorophylls and analogues ³ ; the data include full spectral traces, values for the molar absorption coefficient (ε) and fluorescence quantum yield (Φ_f), and were collected in homogeneous solution at room temperature. All spectral data therein are in digital form and are available to the scientific community.

The availability of spectra is valuable for numerous assessments, including the extent of light‐harvesting, the likely energy of the excited state, the magnitude of electronic coupling in intermolecular energy transfer (e.g., Förster resonance energy transfer, FRET), and for comparative studies across molecules and organisms to understand the interplay of structure and spectra. The absorption spectra are essential for each such assessment. While fluorescence has often been regarded as “the exhaust of the photochemical engine” ⁴ , ⁵ and may be thought to be of little import, knowledge of the fluorescence spectrum is essential for gauging the energy of the singlet excited state, for FRET calculations, and for developing numerous assays and other sensitive detection methods.

The rationale for studies of spectra at low temperature is multifold. The spectral widths are naturally widened by a combination of homogenous broadening (e.g., lifetime, natural line, and collisional broadening) and inhomogeneous broadening (e.g., Doppler thermal broadening, fluctuations of the local environment of the molecule); factors contributing to both are temperature‐dependent. Therefore, the effect of low temperature coupled with immobilizing the chromophore in a solid state (i.e., solvent glass) can significantly diminish spectral broadening. ⁶ , ⁷ Moreover, it is very convenient to have in hand the absorption and fluorescence spectra of photosynthetic chromophores for comparison with those recorded for assemblies in biological matrices, where studies of excitation energy‐transfer and electron‐transfer processes are often carried out at low temperature. Examples of fundamental studies range from (i) the measurements of temperature‐dependent delayed fluorescence as a means of determining the singlet‐triplet energy gap, ⁸ (ii) the use of low‐temperature spectra in the determination of the energy of the weakly allowed Q _x transition in dihydroporphyrin (chlorin) macrocycles, ⁹ and (iii) the evaluation of the equivalence (mirror symmetry) between the absorption and fluorescence spectra. ⁹

Here, we have recorded the low‐temperature (77 K) spectra in frozen glasses of the 12 most common naturally occurring photosynthetic tetrapyrrole macrocycles. In each case, the spectra are compared with those of the same pigment in liquid solution at 293 K. The spectral data in digital form are made publicly available. A strong effort has been made to identify prior publications of such spectra, yet the limitations of present search capabilities have prompted us to record the spectra for the set of compounds. The scope of the search is briefly described. The availability of a systematic assembly of spectral data collected at low temperature should be useful for a variety of experimental and theoretical assessments.

MATERIALS AND METHODS

Pigments were extracted from their native biological sources with 100% methanol (spectroscopic grade). Chls a and b were extracted from parsley leaves. Other pigments were extracted from cells of the following organisms: Chl c ₂ from Symbiodinium, Chl d from Acaryochloris marina, BChl a from Rhodobacter sphaeroides, BChl b from Blastochloris viridis, BChl c from Chlorobaculum tepidum, BChl d from Chlorobium vibroforme, BChl e from Pelodictyon phaeum, and BChl g from Heliobacterium modesticaldum. For spectroscopic measurements, the pigments were purified by high performance liquid chromatography (HPLC) using an Agilent 1100 HPLC system equipped with a Zorbax Eclipse XDB‐C18 reverse phase column (250 mm × 4.6 mm) with an isocratic flow rate of the mobile phase CH₃CN:methanol (MeOH):H₂O (87:10:3, v/v/v) at 1 mL/min. The appropriate pigment‐containing fractions were dried under a stream of nitrogen. The samples were examined immediately following purification by HPLC.

Absorption spectra were recorded using a UV–Vis 1800 spectrophotometer (Shimadzu). Fluorescence spectra were recorded with an RF‐6000 fluorometer (Shimadzu). The spectra were obtained upon excitation of the Q_x band or, if not suitable, the Soret band. The excitation and emission slits were set to afford a bandpass of 5 nm. To minimize scattering and unwanted signal from second‐order diffraction on the spectrometer gratings, adequate long‐pass filters were placed at the detector entrance. For cryogenic measurements, samples were held in the liquid nitrogen vapor‐based VNF‐100 cryostat (Janis Research, USA). All experiments were performed using 1‐cm pathlength cryogenic quartz cuvettes. The spectra were measured with an accuracy of ±0.5 nm, which gives rise to an uncertainty in the Stokes shift of as much as ~30 cm⁻¹. Several solvents are known to afford frozen solutions that are optically transparent, rather than cracking to give an opaque snow. ¹⁰ , ¹¹ , ¹² The samples here were dissolved in spectroscopic grade 2‐methyltetrahydrofuran (2‐MTHF), except for Chl c ₂, for which, due to insolubility in pure 2‐MTHF, a (20:1) (v/v) 2‐MTHF:ethanol mixture was used. Each such solvent, upon cooling to 77 K, formed a transparent glass. The spectra of Chl f and BChl f were collected in 2‐MTHF at 77 K but diethyl ether (DE) at 293 K.

RESULTS

Solvent selection for 77 K measurements

A major challenge in the simultaneous study of many compounds at low temperature is the choice of a medium (solvent) that will be universally suitable for all molecules (no solubility limitations) and, upon cooling to 77 K, will form a transparent glass, which is necessary for recording high‐quality absorption spectra. Moreover, the solvent should minimize or prevent compound oligomerization, which for (bacterio)chlorophylls is a common nuisance and usually manifests by the appearance of a new absorption band at wavelengths longer than the Q _y band of a monomer. Even if negligible at room temperature, aggregation can arise suddenly near the melting point of the solvent glass, given the significant increase in medium viscosity. For example, a study of Chl a in various glass‐forming solvents (EPA—diethyl ether/isopentane/ethanol, 5/5/2, v/v/v; dichloromethane; carbon tetrachloride; benzene; ethanol) ¹³ revealed spontaneous aggregation of the Chl a molecules near the solvent freezing point in most of the selected solvents, even for diluted samples, manifested by substantial changes in the spectral shape of the pigment absorption. Similar problems were also found for Chl a and b dissolved in 2‐methyl pentane. ¹⁴ Therefore, for the current study, we utilized another glass‐forming solvent, 2‐MTHF (freezing point 137 K). All of the compounds studied here, except for Chl c ₂ (which had to be mixed with ethanol), demonstrated very good solubility in 2‐MTHF and marginal or no tendency to self‐aggregation. The process was fully reversible upon thawing, and no sample precipitation was observed. The results were grouped according to the three molecular classes and are provided below.

Absorption and fluorescence spectra of diverse photosynthetic pigments at 77 K

Figure 1 shows spectra at 293 K and 77 K for the porphyrin‐type chlorophyll, Chl c ₂. This compound is utilized by dinoflagellates as an accessory photosynthetic pigment in light‐harvesting (LH) complexes and, due to the widespread ecological appearance of those organisms, is one of the most common chlorophylls in the marine environment. ¹⁵ At 77 K, the absorption spectrum has a typical porphyrin‐like appearance with a dominant Soret band at 468 nm and relatively weak Q _x and Q _y bands (~10% of the Soret band in amplitudes) at 595 and 640 nm, respectively. The fluorescence spectrum at 77 K reveals a narrow (0 → 0) vibronic peak at 642 nm and an exceptionally strong vibronic overtone at 705 nm. Further information on spectral characteristics at 293 K is provided in Table 1. The contents of Table 1 include the peak positions of the chief absorption bands (B, Q_y), the full‐width‐at‐half‐maximum (FWHM) of the Q_y absorption band, the peak positions of the two fluorescence bands, the FWHM of the major fluorescence band, the Stokes shift, and ratios of the intensity of various bands, all in given solvents at 293 or 77 K. In each case, the data from samples in 2‐MTHF at 77 K are displayed in bold font.

FIGURE 1.

Absorption and fluorescence spectra of porphyrin‐type chlorophyll Chl c ₂ at 293 and 77 K in 2‐MTHF. All spectra are normalized at their maxima.

TABLE 1.

Spectroscopic properties of native photosynthetic macrocycles in selected solvents at 293 and 77 K.

Molecule	Absorption					Fluorescence					Solvent	T(K)
	B bands (nm)	Q x (nm)	Q y (nm)	Q y FWHM (cm−1)	Q y/B amp ratio	0 → 0 (nm)	0 → 0 FWHM (cm−1)	Stokes shift, ΔE (cm−1)	0 → 1 (nm)	1/0 amp ratio
Porphyrin chromophore
Chl c 2	468	595	640	380	0.10	642	330	49	705	0.45	2‐MTHF	77
	448	581	629	482	0.10	636	597	174	697	0.38	2‐MTHF	293
	457	587	633	655	0.08	640	570	172	696	0.31	Pyridine	293
	457	581	629	591	0.11						HPLC	293
Chlorin chromophore
Chl a	418, 442		669	268	0.96	672	316	76	729	0.15	2‐MTHF	77
	432		662	392	0.80	669	438	158	724	0.13	2‐MTHF	293
	443		681	429	0.75	680	460	197	738	0.13	Pyridine	293
	430		663	493	0.84						HPLC	293
	432		667	513	0.94						MeOH	293
Chl b	472		653	322	0.40	657	470	93	716	0.26	2‐MTHF	77
	458		646	438	0.36	650	454	95	707	0.17	2‐MTHF	293
	472		655	529	0.33	662	503	161	720	0.18	Pyridine	293
	457		647	562	0.36						HPLC	293
Chl d	402, 463		695	282	1.56	700	385	103	761	0.15	2‐MTHF	77
	395, 450		687	392	1.25	693	436	126	753	0.08	2‐MTHF	293
	406, 464		697	467	1.15	706	477	182	767	0.08	Pyridine	293
	394, 449		690	533	1.16						HPLC	293
Chl f	410, 438		709	380	1.45	724	516	292	794	0.05	2‐MTHF	77
	397, 438		694	467	1.33	707	593	265	776	0.07	DE	293
	413, 443		706	602	1.12	723	642	333			Pyridine	293
BChl c	440		664	181	1.12	665	277	23	723	0.16	2‐MTHF	77
	432		661	336	0.69	665	467	91	721	0.12	2‐MTHF	293
	445		670	400	0.77	675	529	110	732	0.16	Pyridine	293
	432		664	463	0.85						HPLC	293
BChl d	411, 434		653	199	0.91	654	280	23	710	0.15	2‐MTHF	77
	406, 427		650	298	0.83	656	393	141	710	0.21	2‐MTHF	293
	438		659	348	0.71	663	373	92	720	0.10	Pyridine	293
	407, 426		652	425	0.83						HPLC	293
BChl e	475, 489		654	314	0.33	656	420	47	720	0.31	2‐MTHF	77
	462		648	470	0.28	654	507	142	712	0.24	2‐MTHF	293
	476		660	586	0.25	671	745	248	731	0.33	Pyridine	293
	463		651	527	0.27						HPLC	293
BChl f	444, 471		645	291	0.38	647	543	48	708	0.35	2‐MTHF	77
	426, 450		633	329	0.36	643	687	245	698	0.27	DE	293
	469		645	522	0.30	655	794	236	711	0.35	Pyridine	293
Bacteriochlorin chromophore
BChl a	368	600	775	307	2.27	787	321	197	830	0.15	2‐MTHF	77
	363	587	770	513	1.31	780	471	166	830	0.11	2‐MTHF	293
	374	613	781	601	1.31	794	448	210			Pyridine	293
	360	582	770	856	1.09						HPLC	293
BChl b	382, 413	609	801	350	2.13	809	373	123	855	0.07	2‐MTHF	77
	376, 410	596	795	560	1.27	807	516	187	844	0.11	2‐MTHF	293
	385, 418	615	807	647	1.15	825	501	270	874	0.05	Pyridine	293
	372, 407	591	795	846	1.00						HPLC	293
BChl g	371, 410	588	766	397	1.64	774	471	146	835	0.22	2‐MTHF	77
	369, 408	579	765	619	0.92	781	592	268	835	0.20	2‐MTHF	293
	379, 415	597	772	706	0.92	793	601	343	848	0.18	Pyridine	293
	366, 404	568	757	969	0.67						HPLC	293

Note: The bold font indicates data obtained at 77 K.

Four chlorophylls that contain a chlorin‐type macrocycle include Chl a, b, d, and f. While Chl a and b are the most well‐known and common chlorophylls on Earth, the last two representatives are quite rare and in fact were only discovered recently. Chl f was discovered in 2010 in cyanobacterial mats inhabiting stromatolites in Shark Bay, Western Australia. ¹⁶ Chl d was initially reported in 1943 as a minor photosynthetic pigment in red alga, later dismissed as an artifact of the isolation process, and eventually “rediscovered” in the 1990s as a major photosynthetic pigment in the cyanobacterium Acaryochloris marina. ¹⁷ The spectra of Chl a are shown in Figure 2. The notable spectral changes in going from 293 to 77 K include the following: (i) a bathochromic shift of the Soret band from 432 to 442 nm; (ii) a sharpening of the Soret band with greater definition of the band shoulders; (iii) a bathochromic shift of the Q _y band from 662 to 669 nm; (iv) a hyperchromic shift of the Q _y band relative to the Soret band; and (v) a decrease in the Stokes shift, from 158 to 76 cm⁻¹. Aside from substantial spectral narrowing and gain in the vibronic structure within the electronic bands, no side effects such as self‐aggregation (expected absorption band of the aggregate in the 700–710 nm range) were induced upon cooling.

FIGURE 2.

Absorption and fluorescence spectra of chlorin‐type chlorophylls at 293 and 77 K in 2‐MTHF (except Chl f at 293 K in diethyl ether). All spectra are normalized at their maxima.

The spectra of Chl b also are provided in Figure 2. The absorption spectrum at 77 K consists of two major, narrow peaks: the Soret band and the Q _y band at 472 and 653 nm, respectively. The fluorescence peak is at 657 nm. The absorption spectrum at 77 K of Chl d (Figure 2) shows two primary bands, at 438 and 709 nm, and the latter band (Q_y) is dominant. The Soret band, comparable with Chl a, reveals good vibronic resolution with another B band clearly peaking at 402 nm. The fluorescence spectrum at 77 K has a maximum at 700 nm. At 77 K, Chl f (Figure 2) shows a broad Soret band with a dominant peak at 410 nm. The Q _y absorption appears at 709 nm with much higher intensity compared to that of the Soret band. The fluorescence spectrum shows a maximum at 724 nm. Further information on the spectral properties of the chlorin‐type chlorophylls is provided in Table 1.

Figure 3 presents sets of absorption and fluorescence spectra of four chlorin‐type bacteriochlorophylls, namely BChl c, d, e, and f. All are major light‐harvesting pigments exclusively found in the chromosomes of green sulfur bacteria, although BChl f was artificially derived from a BChl e‐producing mutant of the bacterium Chlorobaculum limnaeum and still remains to be found in nature. ¹⁸ , ¹⁹ The spectral features of each member at 77 K are as follows:

• BChl c: The Soret and Q _y bands of BChl c (Figure 3) exhibit maxima at 440 and 664 nm, respectively. The fluorescence spectrum reveals a very small Stokes shift, with the maximum shifted only by 1 nm with respect to the Q _y band (665 nm).

• BChl d: The spectral features (Figure 3) are very reminiscent of those of Chl a; however, the bands are shifted to shorter wavelengths with two clearly resolved B bands (411 and 434 nm) and the Q _y band at 653 nm. Similar to BChl c, the fluorescence spectrum of BChl d exhibits a maximum at 654 nm and demonstrates a very small Stokes shift.

• BChl e: The absorption and fluorescence spectra of BChl e (Figure 3) and BChl f (Figure 3) share a large similarity to those of Chl b. For BChl e, the Soret band consists of clearly resolved peaks (425 and 489 nm), and the Q _y transition appears at 654 nm. The fluorescence spectrum has a maximum at 656 nm, and the spectrum reveals a well‐resolved overtone structure.

• BChl f: The spectral features of BChl f are shifted hypsochromically compared with those of BChl e. The Soret band shows a maximum at 471 nm and has a poorly resolved vibronic structure. The Q _y band appears at 645 nm, while the fluorescence spectrum has a maximum at 647 nm and well‐structured overtone peaks.

FIGURE 3.

Absorption and fluorescence spectra of chlorin‐type bacteriochlorophylls at 293 and 77 K in 2‐MTHF (except BChl f at 293 K in diethyl ether). All spectra are normalized at their maxima.

Note that in the case of BChl c and e, the absorption spectra show some small features at wavelengths beyond the Q _y band at ~700 nm. These are very likely due to spontaneous aggregation of the molecules; however, the small amplitudes render their overall contributions to the absorption spectra essentially negligible. Further information concerning the spectroscopic properties at 293 and 77 K of this class is available in Table 1.

The spectra for the final group of compounds, belonging to the bacteriochlorin chromophore class, are provided in Figure 4. The class consists of three members: BChl a, b present in various purple bacteria, and BChl g found exclusively in heliobacteria. The absorption spectra at 77 K of all three pigments show distinct bands associated with the Soret band (BChl a at 368; BChl b at 382 nm; BChl g at 371 and 410 nm), the Q _x band (at 600, 609, and 588 nm), and the Q _y band (at 775, 801, and 766 nm), respectively. The maxima of the fluorescence spectra are located at 787 nm (BChl a), 809 nm (BChl b), and 774 nm (BChl g). As demonstrated at 77 K, BChl b is the most‐red absorbing and emitting chromophore for the entire (bacterio)chlorophylls group. Additional spectroscopic information concerning these near‐infrared‐active chromophores can be accessed in Table 1.

FIGURE 4.

Absorption and fluorescence spectra of bacteriochlorin‐type bacteriochlorophylls at 293 and 77 K in 2‐MTHF. All spectra are normalized at their maxima.

DISCUSSION

The availability of a systematic set of low‐temperature spectra of a dozen photosynthetic tetrapyrroles should enable a variety of assessments. The material here is organized as follows. First, key changes that are observed in low‐temperature spectra versus those at room temperature are described. Second, several comparisons across the set of compounds are made. Third, the challenges of searching for spectra in the published literature are outlined, including the efforts we made to find low‐temperature spectra of the 12 photosynthetic tetrapyrroles. Fourth, the availability of the data collected herein in PhotochemCAD is briefly discussed.

General changes in spectral features

Low temperature causes noticeable changes in the absorption spectra of organic compounds in condensed media. The magnitude of the change is strongly dependent on the nature of the molecule, the extent and type of molecule–medium interactions, and the composition of the excited state. This is demonstrated in Figure 5, which shows exemplary absorption spectra at 293 and 77 K of BChl a and BChl e. These spectra are of the same solution taken before and after cooling and are given as measured (without normalization). As seen for BChl a, low temperature causes a more than 2‐fold increase in the strength of the Q _y band, although simultaneously the B bands are barely modified. On the contrary, BChl e shows a more uniform response with a comparable increase in the strength of both the Soret and Q _y bands. It is also apparent that low temperature enhances vibrational resolution, which is due to a diminution of the factors that cause spectral broadening (e.g., molecular collisions, Dopler effect, rotational energy).

FIGURE 5.

Temperature‐induced changes in absorbances (in 2‐MTHF) of two exemplary tetrapyrrole macrocycles: BChl a and BChl e.

Comparisons of macrocycles

Some comparisons across the family of macrocycles are possible. Among those with chlorin chromophores, the Stokes shift generally increases with an increasing number and type of conjugated substituents. The conjugated substituents include formyl (F) and vinyl (V). The order of the Stokes shift at 77 K increases as follows (in cm⁻¹): 23 (BChl c, d; 0 V, 0 F), 47 (BChl e; 1 F), 48 (BChl d; 1 F), 76 (Chl a; 1 V), 93 (Chl b; 1 F, 1 V), 103 (Chl d; 1 F), and 292 (Chl f; 1 F, 1 V). It is interesting that BChl c–f, which lack the 13²‐carbomethoxy group, also have the smallest Stokes shift among all of the 12 macrocycles. Considering the other macrocycles, the Stokes shift (in cm⁻¹) for the porphyrin (Chl c ₂) is small (49 cm⁻¹), whereas that for each of the true bacteriochlorophylls is larger (in cm⁻¹): 123 (BChl b), 146 (BChl g), and 197 (BChl a). Finally, the trends at 77 and 293 K are generally similar except for several anomalies. Generally, in 2‐MTHF, the Stokes shift is smaller at 77 versus 293 K in all cases except BChl a (197 vs. 166 cm⁻¹); the values are essentially identical for Chl b (93 vs. 95 cm⁻¹). [Note that the spectra of Chl f and BChl f were collected in 2‐MTHF at 77 K but in diethyl ether at 293 K; insufficient sample precluded more exacting experimentation.] All other things being equal, the smaller Stokes shift would support more extensive homo‐transfer of excited‐state energy in the weak coupling limit. ²⁰ Whether the differing values of the Stokes shift bear on evolutionary selection of a given chromophore for a biological function, or are only spectroscopic tidbits of academic interest, is a consideration outside the scope of the present paper.

Another interesting feature of the macrocycles is the ratio of the intensity of the long‐wavelength (Q _y) band to that of the Soret (B) band. The ratio is small (0.10) at 293 and 77 K for the porphyrin (Chl c ₂) but increases at the lower temperature for all other macrocycles examined here. The increase is slight (1.1‐fold, Chl b) to substantial (~1.7 fold, BChl a, b). In the most pronounced cases, which occur for BChl a and BChl b, the Q _y/B ratio is 2.27 and 2.13, respectively.

Literature data

Finding relevant sources for absorption and fluorescence spectra of (bacterio)chlorophylls in the literature is a surprisingly challenging task. A discussion of this topic is germane as it is emblematic of the challenges to conduct scholarly analyses of chromophores in the field of photochemistry and photobiology. The vast size of the literature—estimated at >10⁸ papers ²¹ —is perhaps counterintuitively no longer the intrinsic problem, as modern search engines can readily canvass such a large collection. The challenge in finding spectra and photophysical data is that existing search engines have limited capacity to identify graphical images, to reliably extract numerical values, and to affiliate any extracted information with valuable experimental context. We recently found that chatbots may provide some utility in this regard but hardly are a panacea. ²²

The search process complicates even further for low‐temperature data. The literature reports of low‐temperature spectra of (bacterio)chlorophylls (either absorption and fluorescence) are scattered and tend to display spectra for one compound in a given frozen solution (i.e., glass). Systematic collections of a set of those molecules simultaneously studied in a selected medium and at a specific temperature (77 K, 4.5 K, etc.) do not appear to be available. Moreover, even finding relevant sources based on a reasonable string of keywords can be extremely challenging. For example, use of the phrase “low‐temperature absorption of chlorophyll” in Google Scholar returned ~629,000 results; the more specific expression “low‐temperature absorption of chlorophyll‐a” still gave an astonishing ~319,000 results for this specific pigment.

Considering stricter search engines such as Clarivate's Web of Science (search details: Web of Science Core Collection, all fields), application of the original phrase “low‐temperature absorption of chlorophyll” (omitting the “a”) gave 951 results, and narrowing the search toward Chl a (second phrase) returned 441 results. On the contrary, the key phrase “low‐temperature absorption of bacteriochlorophyll” returned 264 results, clearly reflecting narrowing specificity of keywords and lesser research on the more architecturally complex pigment of less extensive ecological reach. It warrants mention that Google Scholar performs a meta search on entire documents whereas Web of Science searches only the Abstracts and Keywords.

Quite surprising outcomes were observed upon performing searches directed toward fluorescence. Searches concerning the phrase “low‐temperature fluorescence of bacteriochlorophyll” returned only 135 potentially relevant results, although the same phrase for “chlorophyll” afforded 4389 results. Such disparity was clearly associated with many studies of chlorophyll fluorescence done for whole plant cells. A manual search to find spectra in papers numbering on the order of 10²–10³ appears far more manageable than the seemingly impossible task of sifting through ~10⁵ published articles. It warrants emphasis, however, that the search process can unintentionally narrow because alternative terms may be used to denote “low temperature” that are not captured in a search for this term; examples may include “cryogenic temperature” or “77 K.” Thus, the focused search space for the topic at hand may not be merely 10²–10³ papers.

The abundance of search results is not spread equally over the set of 12 tetrapyrrole macrocycles. The recently discovered molecules, or those from rare sources, have naturally received less attention, and the results in those cases are expected to be very limited. Here, the 951 papers identified from the Web of Science search for “low‐temperature absorption of chlorophyll” were examined manually by one of us (D.M.N.). Despite the initially vast search outcome, an extensive record of pertinent low‐temperature spectra of chlorophylls in solution was not found; indeed, most of the papers so‐identified were devoid of the desired data. The papers were limited in the following ways: (i) Chlorophyll was described albeit in biological matrices, (ii) the terms “low temperature” and “absorption” and “chlorophyll” were present but unconnected, and/or (iii) low temperature referred to studies at physiologically low temperatures, such as 4°C, not 77 K or in a solvent glass. As expected, the most known chlorophyll, Chl a, revealed the largest spectral record (absorption ⁹ , ¹³ , ¹⁴ , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ and fluorescence ⁹ , ¹⁴ , ²⁶ , ³¹ , ³² , ³⁴ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ ). On the contrary, we could not find any low‐temperature absorption spectra of BChl g and BChl b, although fluorescence spectra at 77 K of both were recently published. ⁴³ Only a few results were found for Chl d (absorption ⁴⁵ , ⁴⁶ and fluorescence ³⁶ , ⁴³ , ⁴⁷ ), Chl f (absorption and fluorescence ⁴⁸ , ⁴⁹ ), Chl c ₂ (absorption ²⁵ and fluorescence ⁴³ ), and BChl f (absorption and fluorescence ⁵⁰ ), and those spanned temperatures from 1.4 K to 183 K and were measured in glasses of various solvents.

A surprisingly limited record was also identified for the more common chlorin‐type bacteriochlorophylls: BChl c, d, and e. Attempts to find low‐temperature absorption spectra in any glass‐forming solvents were nearly fruitless (BChl c ⁵¹ , ⁵² ), although absorption spectra at 77 K of these three compounds in chromosomes are available. ⁵³ , ⁵⁴ On the contrary, low‐temperature fluorescence spectra of BChl c, d, and e were published (BChl c, ⁴³ , ⁵² BChl d, ⁴³ BChl e ⁴³ ). One would also expect that another common chlorophyll, Chl b, would be broadly covered in the literature. Surprisingly, however, finding low‐temperature absorption or fluorescence spectra in any kind of glass for Chl b was very difficult. For absorption, the most relevant search led to an absorption spectrum at 10 K of the pigment reconstituted in a non‐native protein complex, ⁵⁵ and spectra at 77 K could be found in work focused on chlorophyll aggregation. ¹⁴ On the contrary, fluorescence spectra at 77 K could be found in more sources, ¹⁴ , ³⁸ , ⁴⁰ , ⁴³ sometimes buried in the work on very different subjects, as, for example, in a study of the phosphorescence of (bacterio)chlorophylls. ⁴³ Nor does BChl a, the most common bacteriochlorin‐type pigment, have an extensive record of low‐temperature spectra in solvents: relatively few absorption ³⁷ , ⁵⁶ , ⁵⁷ , ⁵⁸ and fluorescence ³⁷ , ⁴³ , ⁵⁶ , ⁵⁷ , ⁵⁸ spectra were found. In this case, the vast amount of low‐temperature spectra that exist in the literature is for BChl a bound to natural matrices, primarily in light‐harvesting antenna complexes from various species of purple bacteria (LH2 and LH1) or Fenna‐Matthews‐Olson (FMO) complexes from green sulfur bacteria.

All taken together, there appears to be a large quantity of published low‐temperature spectra of various photosynthetic macrocycles. Such data often concern pigments in some biological matrix such as light‐harvesting complexes, reaction centers, or whole cells. If the spectra of interest are not clearly indicated in the publications, like a single needle in an immense haystack, there is limited likelihood of finding them other than by happenstance.

Data availability

The availability of data is essential for scientific progress. Over the years, we have been developing a software program and accompanying spectral databases for use in photochemistry, photobiology, photophysics, and photonics. The development of PhotochemCAD was motivated by the desire to have at one's fingertips a vast collection of curated spectral data for a wide variety of representative compounds and the ability to perform photophysical calculations using such spectral data. ⁵⁹ , ⁶⁰ The spectral data typically include absorption spectra with molar absorption coefficient values, fluorescence spectra with quantum yield values, and references to the original photochemical literature. Three generations of software development and revision have been carried out since the zeroth version created in the 1980s in pursuit of this vision. ⁵⁹ , ⁶¹ , ⁶² , ⁶³ , ⁶⁴ Numerous spectral databases have been assembled, including those of tetrapyrrole macrocycles, which to date encompass naturally derived porphyrins, ⁶⁵ chlorophylls and analogues, ³ tolyporphins, ⁶⁶ and synthetic chlorophylls. ⁶⁷ All of the spectral databases assembled thus far incorporate solution spectra acquired at room temperature. The low‐temperature spectra collected herein will be made available at the PhotochemCAD site.

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

DATA AVAILABILITY STATEMENT

All data described in this paper are contained here and, moreover, are available for downloading at no cost at the website www.photochemcad.com.