A database for deliquescence and efflorescence relative humidities of compounds with atmospheric relevance

Deliquescence relative humidity (DRH) and efflorescence relative humidity (ERH), the two parameters that regulate phase state and hygroscopicity of substances, play important roles in atmospheric science and many other fields. A large number of experimental studies have measured the DRH and ERH values of compounds with atmospheric relevance, but these values have not yet been summarized in a comprehensive manner. In this work, we develop for the first-of-its-kind a comprehensive database which compiles the DRH and ERH values of 110 compounds (68 inorganics and 42 organics) measured in previous studies, provide the preferred DRH and ERH values at 298 K for these compounds, and discuss the effects of a few key factors (e.g., temperature and particle size) on the measured DRH and ERH values. In addition, we outline future work that will broaden the scope of this database and enhance its accessibility.


Introduction
Hygroscopicity is one of the most important physicochemical properties of substances as it regulates their interactions with water vapor under sub-and super-saturated conditions [1][2][3] .The hygroscopicity of aerosol particles is of critical importance in atmospheric science due to its effects on the environment and climate.Under a subsaturated condition, such as when the relative humidity (RH) is < 100%, aerosol particles absorb or adsorb water vapor from the surrounding environment due to hygroscopicity, leading to changes in particle size and mass and thus affecting their optical and radiative properties [ 1 , 4 , 5 ].Under a supersaturated condition (RH > 100%), hygroscopicity is closely related to cloud condensation [ 1-3 , 6 , 7 ] and ice nucleation activities [ 3 , 8-11 ] of aerosol particles, thereby influencing their potential to be activated to cloud droplets and ice crystals.Furthermore, the liquid water contents of aerosol particles will increase with increasing RH due to hygroscopicity, causing changes in other important physicochemical properties of aerosol particles, such as the phase state [12][13][14][15] , acidity [ 16 , 17 ], and heterogeneous and multiphase reactivity [18][19][20][21] .
Hygroscopicity is also of great interest in many other fields [22] , such as thermodynamics, chemical engineering, food and pharmaceutical science, and earth and space science.It determines phase transi-tions (deliquescence and efflorescence) of hygroscopic compounds at different RHs [12] , and water activities and thermodynamics of aqueous solutions are directly related to the hygroscopic properties of solutes [23] .NaCl, a widely used salt in chlor-alkali industry, is crystalline at < 75% RH and will be transformed to an aqueous electrolyte solution at > 75% RH.In addition, its hygroscopicity significantly impacts the production of downstream chemicals, such as NaOH, H 2 and Cl 2 [24][25][26] .Hygroscopicity also affects storage and transportation in the food industry.For example, deliquescence will accelerate the degradation of labile food ingredients and decrease the stability of powdered food [ 27 , 28 ].It plays key roles in the chemical and physical stability of pharmaceuticals, and water uptake by drug powders significantly determines their quality, safety, and efficacy [29][30][31][32][33][34][35][36][37][38] .Furthermore, in atomizing inhalation treatments, transport and deposition of pharmaceutical ingredients in the human upper airway and lung are closely related to the hygroscopicity of drug aerosols [ 32 , 35 , 37 , 39-46 ].It is currently believed that one prerequisite for habitability is the existence of liquid water [ 47 , 48 ].Although pure liquid water is thermodynamically unstable in the hyper arid environments found on Earth, Mars and probably other planets, highly hygroscopic materials such as chlorates and perchlorates can take up water even at RHs significantly lower than 100%, leading to the formation of aqueous solutions and liquid water [49][50][51][52][53][54][55] .Inorganic salts, such as (NH 4 ) 2 SO 4 , are solid at very low RHs.As RH increases, a particle remains solid until it reaches a specific value ( ∼80% for (NH 4 ) 2 SO 4 at 298 K) at which the solid particle will abruptly take up large amounts of water vapor and be transformed to a saturated droplet ( Fig. 1 ).This solid-to-liquid phase transition is called deliquescence, and deliquescence relative humidity (DRH) is defined as the RH at which deliquescence takes place.As RH further increases, the deliquesced particle will take up more water vapor, in order to maintain thermodynamic equilibrium between water vapor and aqueous water in the solution.
In contrast, water gradually evaporates from the aqueous particle as RH decreases.However, the aqueous particle will not be transformed to a solid particle when RH is decreased to DRH, and instead it will exist as a supersaturated droplet.This is because supersaturation is essential for an aqueous particle to overcome the excess energy barrier (larger than the Gibbs free energy of a crystalline solid) to form nucleation germs that will induce homogeneous nucleation and efflorescence transitions.When the RH is reduced to a specific RH that is lower than the DRH, efflorescence will occur and the supersaturated droplet will be transformed to a solid particle.The RH at which efflorescence occurs is called the efflorescence relative humidity (ERH).Hysteresis between deliquescence and efflorescence has been observed for many compounds, and the phase state of an aerosol particle is determined not only by the current RH but also by the RH history it has experienced.
It should be emphasized that not all the single-component particles exhibit well-defined deliquescence and efflorescence ( Fig. 1 ).For example, sulfuric acid (H 2 SO 4 ), an extremely hygroscopic material, takes up or losses water continuously as RH changes without deliquescence or efflorescence.As a result, H 2 SO 4 particles always stay in a liquid state.NH 4 HSO 4 displays a different phase transition when compared to (NH 4 ) 2 SO 4 and H 2 SO 4 : solid NH 4 HSO 4 particles will be deliquesced to form droplets when RH is increased to 30-41%, whereas NH 4 HSO 4 droplets will not effloresce to form solid particles even when RH becomes very low (down to 10%).Therefore, once deliquesced, NH 4 HSO 4 particles will remain as aqueous droplets.
The important roles of databases that summarize thermodynamic and kinetic data have been well recognized in atmospheric chemistry [ 57 , 58 ].Currently databases are available for kinetic and photochemical data [59-61] , Henry's law constants [62] , gas phase diffusion coefficients [63][64][65] , and etc.Despite the importance of DRH and ERH in many fields, up to now these values have not yet been compiled or summarized in a comprehensive manner.Martin [12] discussed phase transition processes of atmospheric particles, and summarized DRH and ERH values of selected species ( < 20 compounds) at 298 K. Very recently, Ma et al. [66] reviewed efflorescence kinetics of atmospherically relevant particles, and summarized the measured ERH for seven inorganic compounds, including NaCl, KCl, (NH 4 ) 2 SO 4 , Na 2 SO 4 , MgSO 4 , NaNO 3 , and NH 4 NO 3 .
In this work, we have developed for the first time a comprehensive database in which DRH and ERH values measured in previous studies are compiled.This database summarizes DRH and ERH of a total of 110 compounds, including 68 inorganics and 42 organics.It is mainly focused on compounds with atmospheric relevance, but also includes certain inorganic compounds relevant for earth and planetary sciences, such as perchlorates [ 49 , 50 , 53 , 54 ].The motivations for the development of such a database are (1) to provide a comprehensive summary of measured DRH and ERH values and (2) to identify knowledge gaps that will stimulate measurements of DRH and ERH values in future work.In addition, many studies which measured DRH or ERH values of some compounds also reported their hygroscopic growth factors (or waterto-solute ratios) as a function of RH (or water activities).Although our present work does not compile growth factors or water-to-solute ratios for compounds included, it can serve as a starting point (i.e., as a collection of references) for people who are interested in growth factors or water-to-solute ratios measured by previous studies.

Techniques used to measure deliquescence and efflorescence relative humidities
A large number of experimental techniques have been employed to measure DRH and ERH (and hygroscopicity in general) of compounds with atmospheric relevance.Samples under investigation broadly include four types, such as bulk solutions, samples deposited on substrates, levitated single particles, and aerosol particles [22] .These techniques were reviewed and discussed in a recent paper [22] , and therefore, here we only describe in brief four techniques widely used to measure DRH and ERH.
The nonisopiestic method is a bulk solution-based technique and measures RH of the air in equilibrium with an electrolyte solution with a given concentration (i.e., a given water-to-solute ratio), and DRH is equal to the RH of the air in equilibrium with the saturated solution.This method has been used to measure DRH values of many electrolytes and their water-to-solute ratios as a function of water activity [67][68][69][70] .For example, it was used to measure DRH of 28 electrolytes (including sulfates, nitrates and halides) as a function of temperature (273-373 K) [70] , and reported DRH values ranged from ∼3% to 98%.This method is simple and accurate, but it cannot be used to determine ERH values.
Hygroscopicity can be quantified by measuring sample mass as a function of RH, using a RH-controlled balance [ 71 , 72 ] or a vapor sorption analyzer [ 73 , 74 ], and DRH is equal to the RH at which an abrupt increase in sample mass is observed during humidification.For example, a vapor sorption analyzer was used to determine DRH values of six inorganic compounds (CaBr 2 , MgCl 2 •6H 2 O, Mg(NO 3 ) 2 •6H 2 O, NaCl, (NH 4 ) 2 SO 4 and KCl) at different temperatures (278-303 K) [74] , and the measured values ( < 20% to > 85%) agreed well with those reported by Greenspan [70] .Similar to the nonisopiestic method, this method cannot be used to determine ERH values either.
Electrodynamic balance (EDB) is a single-particle levitation technique widely employed to study aerosol hygroscopicity [75][76][77] .In a typical EDB measurement, a single particle (typically 1-100 m in diameter) can be levitated in the EDB chamber by adjusting the alternating current (AC) and direct current (DC) electric fields surrounding the particle, and relative mass change of the particle (for example, due to condensation or evaporation of water) is equal to the change in the DC voltage needed to levitate the particle.The temperature and RH in the EDB chamber can be regulated, and the RH at which an abrupt increase (or decrease) in the relative mass of the sample occurs during humidification (or dehumidification) is equal to its DRH (or ERH).Tang and Munkelwitz [77] employed this technique to determine the DRH values of six inorganic salts ((NH 4 ) 2 SO 4 , Na 2 SO 4 , NaNO 3 , NH 4 NO 3 , KCl and NaCl) as a function of temperature (278-308 K), and found that the temperature dependence of their DRH values could be well described by the Clausius-Clapeyron equation.In addition, EDB has been used to measure ERH values of many compounds with atmospheric relevance [78][79][80] .
The humidity tandem differential mobility analyzer (H-TDMA), which measures the mobility diameters of aerosol particles at different RHs, has been extensively used to investigate hygroscopic properties of aerosol particles [81][82][83][84] .The measured DRH (or ERH) is defined as the RH at which a sudden increase (or decrease) in aerosol mobility diameter takes place during humidification (or dehumidification).An H-TDMA was employed to measure DRH values and hygroscopic properties of 10 water-soluble carboxylic salts at 293 K [85] : the DRH was determined to be 39-42% for sodium acetate, 81-82% for sodium pyruvate, and > 90% for sodium oxalate and ammonium oxalate, while the other six salts (sodium malonate, sodium succinate, sodium maleate, ammonium tartrate, sodium tartrate and humic acid sodium salts) displayed continuous water uptake when RH increased from 5% to 90%.This technique has also been widely used to measure ERH values of particles with atmospheric relevance [86][87][88][89][90][91][92] .

Overview of the database
The database we have developed provides a comprehensive compilation of DRH and ERH of 110 compounds measured at different temperatures in previous work.These species were classified into 11 groups of inorganic species (including six sulfates, three bisulfates, nine nitrates, three fluorides, 12 chlorides, eight bromides, six iodides, four chlorates, three iodates, 10 perchlorates and four carbonates) and four groups of organic species (including five methanesulfonates, 12 monocarboxylic salts, 13 dicarboxylic acids and 12 dicarboxylic salts).For each compound, their DRH and ERH values measured by previous studies at different temperatures are summarized in individual tables and also briefly discussed (in supplementary materials).This document to a large extent adopts the format used by the International Union of Pure and Applied Chemistry Task Group on Atmospheric Chemical Kinetic Data Evaluation ( https://iupac-aeris.ipsl.fr/).In addition, we have attempted to provide the preferred DRH and ERH values at 298 K for compounds included in the database, and these values are summarized in Table 1 .
As shown in The interaction of water vapor with compounds of atmospheric relevance has been much less investigated during dehumidification than humidification, and the number of compounds for which the ERH values have been measured is much smaller than the number of compounds for which the DRH values have been measured.This is the major reason why we were able to provide preferred ERH values at 298 K for only 40 compounds where n is the solubility of the solute in water (moles of solute per mole of water), ΔH s is the enthalpy of dissolution (J mol − 1 ), and ΔH v is the heat of water vaporization (J mol − 1 ).As the Clausius-Clapeyron equation can be used to describe vapor pressures of pure water and over the solution at different temperatures, the dependence of DRH on temperature can be obtained: where T is the temperature (K) and R is the universal gas constant (8.314J mol − 1 K − 1 ).If we assume that ΔH s , the enthalpy of dissolution, does not vary with temperature, the integration of Eq. 2 gives Eq. 3 : where DRH( T ) and DRH (298) are the DRH values at T and 298 K, respectively.The three parameters, A, B and C , describe the dependence of solubility ( n ) on temperature, as expressed by Eq. 4 : If we further assume constant solubility (i.e., B = C = 0), Eq. 3 can be simplified to Eq. 5 , which is widely used to approximate the dependence of DRH on temperature due to its simplicity.
Further information on how to derive Eqs. 3 and 5 can be found elsewhere [ 77 , 93 , 94 ].The temperature dependence of DRH is determined by the overall enthalpy change of deliquescence ( ΔH ).DRH decreases with temperature if ΔH is positive and increases with temperature if ΔH is negative.
It shows DRH values of NH 4 NO 3 , KCl, NaCl and CH 3 COONa at different temperatures measured by previous studies (Fig. 2) .As displayed in Fig. 2 a, the DRH values of NH 4 NO 3 show a negative dependence on temperature, decreasing from 79 ± 1% at 274 K to 48.4% at 323 K, and the temperature dependence can be well approximated by Eq. 5 .Similarly, the DRH values of KCl also decrease with temperature, from 88.6 ± 0.5% at 273 K to 78.5 ± 1.0% at 363 K. Nevertheless, as shown in Fig. 2 b, Eq. 5 can only describe the temperature dependence of its DRH values for temperature between 283 and 323 K, while the DRH values measured at 328-363 K [70] are significantly larger than those extrapolated from the fitted curve using Eq. 5 .This underscores that extrapolation using Eq. 5 may not always be valid and should be conducted with caution.
The DRH values were determined to be 73-84% at 253-353 K for NaCl ( Fig. 2 c), showing no obvious dependence on temperature.In addition, the DRH values showed a positive temperature dependence for CH 3 COONa ( Fig. 2 d), increasing from 27-29% at 270 K to 43-45% at 298 K, and the positive dependence can be well approximated using Eq. 5 .

Hydration form
It was argued in a previous study [104] that DRH values would be different for anhydrous and hydrated forms of the same compound, and this was supported by some previous measurements.For example, the DRH values at 223-253 K were found to be < 20% for CaCl 2 •2H 2 O and 50-80% for CaCl 2 •6H 2 O [105] .However, one may argue from a thermodynamic view that DRH values should be identical for anhydrous and hydrated forms.Take MgCl 2 and MgCl 2 •6H 2 O as an example: once deliquesced, MgCl 2 and MgCl 2 •6H 2 O will both be transformed to saturated MgCl 2 solutions with identical water activities; as a result, MgCl 2 and MgCl 2 •6H 2 O should have the same DRH value at a given temperature.This was also supported by some previous experimental work.For example, good agreement in DRH values was found between MgCl 2 [ 69 , 70 ] and MgCl 2 •6H 2 O [ 74 , 106 ], and between Ca(NO 3 ) 2 [ 69 , 107 ] and Ca(NO 3 ) 2 •4H 2 O [ 67 , 106 , 108 ].More experimental and theoretical work is required to better understand the effects of hydration forms on DRH values.

Particle size
Several previous studies measured DRH values as a function of particle size [ 86 , 87 , 91 , 109-116 ].The RH on the curved surface of a droplet can be described by the Köhler theory [ 117 , 118 ]: where a w is the water activity in the solution, K e is the Kelvin term which describes the increase in vapor pressure over a curved surface relative to that over a flat surface, M w is the molar mass of water (kg mol − 1 ),  is the surface tension of the solution (J m − 2 ),  w is the density of water (kg m − 3 ), and d p is the droplet diameter (m).The Kelvin effect is negligible for particles larger than 100 nm [ 3 , 118 ], and this is supported by measurements that found no significant difference in the DRH values between supermicrometer ( > 1000 nm) and submicrometer (between 100 and 1000 nm) particles.However, for multiple component particles, their morphology and mixing states may vary with particle size, leading to variations in the DRH values (and also hygrosocpic properties) with particle size [113] .
The Kelvin effect becomes important and particle size plays an indispensable role in DRH for particles smaller than 100 nm.Fig. 3 displays the DRH values as a function of particle size (6-60 nm) at 298 K for (NH 4 ) 2 SO 4 , NaCl, KCl, KBr and KI.It is evident that the DRH values show a negative dependence on particle size for all four compounds.For example, the DRH values of NaCl at 298 K decrease from 87 ± 2.5% at 6 nm to 76 ± 2.5% at 60 nm [86] , and the dependence of DRH on particle size can be well described by Eq. 6 .We also note that a differential Köhler analysis coupled with the Ostwald-Freudlich equation [112] can also describe the size dependence of DRH values.

Initial phase states
The occurrence of deliquescence is closely related to the initial phase state of particles [90] .For example, the DRH values at 298 K were measured to be approximately 49.5%, 52.5%, 31.5%, and 28.5% for crystalline Ca(NO 3 ) 2 •H 2 O, Mg(NO 3 ) 2 •6H 2 O, MgCl 2 •6H 2 O and CaCl 2 •6H 2 O, respectively, while continuous hygroscopic growth was observed for Ca(NO 3 ) 2 , Mg(NO 3 ) 2 , MgCl 2 and CaCl 2 aerosol particles that were amorphous [106] .The DRH of oxalic acid at around room temperature has been widely investigated [ 90 , 119-125 ], and different deliquescence transitions were observed: DRH was measured to be > 90% RH in five studies [119][120][121][122][123] , while three other studies [ 90 , 124 , 125 ] suggested continuous water uptake with increase in RH.This discrepancy is attributed to the difference in initial phase states of oxalic acid particles (crystalline versus amorphous), which results from different particle diffusion drying methods.Different pretreatments of aerosol particles which were generated via atomization could also affect the measured  DRH values.For example, after atomized KNO 3 particles were heated overnight at 104 °C, they would be transformed to crystalline solids and become deliquesced at > 90% RH [126] ; for comparison, continuous water uptake was observed for atomized KNO 3 particles without heating.

Substrates used to support samples
Some studies [ 116 , 127 ] examined whether substrates used to support samples would affect the measured DRH values, and they found their effects to be negligible.For example, Eom et al. [127] investigated the influence of six supporting substrates on the measured DRH values of (NH 4 ) 2 SO 4 , NaCl, and KCl particles using optical microscopy.
The measured DRH values agreed well with theoretical values for particles deposited on hydrophobic substrates (transmission electron microscopy grids and parafilm-M); while the measured DRH values were 1-2% lower than the theoretical values for particles deposited on hydrophilic substrates (Al foil, Ag foil, silicon wafer and cover glass).The authors suggested that the observed lower DRHs were attributed to a slight increase in the Gibbs free energy for solid particles due to absorbed water on the particles and hydrophilic substrates prior to deliquescence [127] .However, such small differences might be insignificant.

Temperature
Four previous studies [128][129][130][131] measured ERH values of (NH 4 ) 2 SO 4 at different temperatures.As shown in Fig. 4 a, Xu et al. [128] found that it first decreased with temperature from 54% at 254 K to 37% at 283 K and did not change significantly when the temperature was further increased to 308 K; Cziczo and Abbatt [129] found that it decreased with temperature from 41 ± 6% at 238 K to 37 ± 3% at 273 K; Onasch et al. [130] suggested that it first decreased with temperature from 39.0 ± 6% at 234 K to 30.5 ± 2.5% at 273 K, and then slightly increased with temperature to 32.0 ± 1% at 295 K.Although these three studies [128][129][130] showed substantial discrepancies, one may conclude that ERH values of (NH 4 ) 2 SO 4 first decreased with temperature (up to ∼280 K) and then did not change significantly when the temperature was further increased.However, the fourth study [131] suggested that it increased slightly with temperature from 28.5 ± 2.5% at 260 K to 30.8 ± 2.5% at 263.5 K.The ERH values also showed a negative temperature dependence for CH 3 SO 3 Na ( Fig. 4 b), decreasing from 63-65% at 268 K to 50-52% at 296 K.
The ERH values of NaCl were measured at different temperatures by two previous studies [ 99 , 100 ] that both revealed positive dependence on temperature.As shown in Fig. 4 c, ERH values of NaCl were reported to increase with temperature from 35 ± 4% at 253 K to 45 ± 3% at 283 K [99] and from 43 ± 2% at 268 K to 48 ± 2% at 296 K [100] .Gao et al. [103] measured ERH values of sodium formate, sodium acetate, and sodium succinate at different temperatures, and they revealed positive dependence of ERH values for the three compounds, similar to NaCl.As shown in Fig. 4 d, ERH values were found to increase with temperature from 16-21% at 270 K to 31-37% at 296 K for sodium formate, 17-21% at 270 K to 39-43% at 296 K for sodium acetate, and 34-38% at 270 K to 50-57% at 296 K for sodium succinate.
Compared to DRH values, the temperature dependence of ERH values has been less investigated.It was revealed by previous studies that ERH values could decrease, increase, or show no significant change with an increase in temperature, similar to DRH values.To the best of our knowledge, there are no equations that can generally describe or approximate the dependence of ERH values on temperature, partially because relevant experimental studies are very limited, and measured ERH values are more uncertain when compared to measured DRH values.

Particle size
Classical nucleation theory suggests that the expectation time ( ) for production of a crystalline germ at a specific temperature or relative humidity can be described by Eq. 8 [12] : where J is the homogeneous nucleation rate (cm − 3 s − 1 ) and V is the aqueous solution droplet volume (cm 3 ).Eq. 8 suggests that the expectation time for a single nucleation germ formation requires less time for larger particles.Therefore, nucleation germ formation is more favored for larger particles [12] , and consequently the ERH values are expected to increase with particle size.Laskina et al. [113] found that the ERH values of (NH 4 ) 2 SO 4 and NaCl at 298 K both increase with particle size: the ERH of (NH 4 ) 2 SO 4 was measured to be 37.7 ± 4.1% for 100 nm particles and 43.5 ± 2.1% for 3-10 m particles, and the ERH of NaCl was measured to be 43.0 ± 1.0% for 100 nm particles and 52.8 ± 1.1% for 3-10 m particles.In contrast, ERH values were determined to be 35-40% for 100 nm and 5-25 m (NH 4 ) 2 SO 4 particles [89] , showing no dependence on particle size.
To summarize, although it is theoretically suggested that ERH values would increase with particle size, different size dependences of ERH values were observed for different compounds.To better understand the size dependence of ERH values, further experimental measurements are encouraged, and one critical aspect of such measurements should be to reduce the uncertainties of measured ERH values.

Substrates used to support samples
The contact of supporting substrates with aqueous particles can significantly influence their nucleation mechanisms [12] .For airborne or levitated aqueous particles, critical germ formation occurs via homogeneous nucleation.For droplets deposited on substrates, the critical cluster formation frequency increases at the substrate interface, thus the formed crystal germ at the interface can induce heterogeneous nucleation of droplets, leading to efflorescence at higher RHs.
However, not all the substrates on which aqueous droplets are deposited will promote efflorescence transitions.For example, Eom et al. [127] suggested that the observed ERH values of (NH 4 ) 2 SO 4 , NaCl and KCl droplets deposited on transmission electron microscopy grids, parafilm-M, Al foil, silicon wafer and cover glass agreed well with those for airborne or levitated particles.In another study [141] , the ERH was determined to be ∼45% at 298 K for NaNO 3 particles deposited on CaF 2 crystals, in good agreement with those (35-45%) reported by other studies [ 139 , 140 , 142 ].

Summary and outlook
Deliquescence and efflorescence are two phase transition processes of critical importance in atmospheric science and many other fields.A large number of experimental studies have been carried out to measure deliquescence relative humidity (DRH) and efflorescence relative humidity (ERH) of compounds with atmospheric relevance.However, DRH and ERH values have not yet been summarized in a comprehensive manner.In this work, we have developed a comprehensive database that summarizes the measured DRH and ERH values of 110 compounds (68 inorganics and 42 organics), and provided the preferred DRH and ERH values at 298 K for these compounds ( Table 1 ).In addition, we have also discussed the effects of a few key factors (e.g., temperature and particle size) on the measured DRH and ERH values.
A comprehensive summary of DRH and ERH values has been provided in the supplementary materials.We also plan to upload our summary sheets to a website that is funded by the National Natural Science Foundation of China (this website is under construction and will become online in 2022).This website will have an interactive and userfriendly interface in order to increase data accessibility, and it will also enable dynamic updates to include additional data in time.Currently this database only includes 42 organic compounds (i.e., five methanesulfonates, 12 monocarboxylic salts, 13 dicarboxylic acids, and 12 dicarboxylic salts), while there are many thousands of organic compounds contained by atmospheric aerosols.One major task in the future is to expand this database to include additional organic compounds with atmospheric and pharmaceutical relevance.
For the 110 compounds included in the database, 99 compounds have preferred DRH values and 40 compounds have ERH values ( Table 1 ) at 298 K.One primary reason that the DRH and ERH values are not provided for many compounds is that large discrepancies found for reported DRH and ERH values precluded us from providing the preferred values.Therefore, careful and robust measurements are needed to reduce the uncertainties in the DRH and especially the ERH values of these compounds.
For many compounds included in the database, DRH and ERH measurements have only been provided at around room temperature ( ∼298 K), and the temperature dependence of DRH and ERH was only examined for a limited number of compounds.The temperature in the troposphere can range from < 200 to > 300 K; as a result, DRH and ERH measurements at different temperatures (especially < 273 K) will be very useful to better understand phase transitions in the troposphere as well as under other cold environments (such as Mars) and to elucidate temperature dependence of DRH and ERH at the fundamental level in physical chemistry.
DRH and ERH are two RH thresholds at which abrupt changes in liquid water contents take place due to an increase or decrease in RH.In addition to DRH and ERH values, it is very important to know changes in the liquid water contents (i.e., hygroscopic growth factors or water-tosolute ratios) with RH.While the current database can serve as a starting point to summarize hygroscopic growth factors or water-to-solute ratios, we have not included these experimental data at this moment as it re-

Fig. 1 .
Fig. 1.Relative volume change of (NH 4 ) 2 SO 4 , NH 4 HSO 4 and H 2 SO 4 particles at 298 K as a function of RH .V is the particle volume at a given RH and V 0 is the volume of the dry particle.The data used in the figure were produced using the extended aerosol inorganics model (E-AIM, http://www.aim.env.uea.ac.uk/aim/aim.php) [56] .

3. 1 .
Deliquescence relative humidity 3.1.1.TemperatureDeliquescence consists of two processes: condensation of water vapor on the solute followed by dissolution of the solute.The overall enthalpy change of the deliquescence ( ΔH ) can be expressed by Eq. 1 :

Table 1 ,
the preferred DRH values at 298 K are provided for 99 compounds among the 110 compounds included in the database.Such values are not provided for the other 11 compounds, including MgSO 4 , NaHSO 4 , KHSO 4 , Ca(NO 3 ) 2 , LiClO 4 , Ca(ClO 4 ) 2 , CH 3 SO 3 NH 4 , Mg(CH 3 SO 3 ) 2 , sodium malonate, sodium maleate, and sodium tartrate.This is due to one or more of the following reasons: 1) no DRH values have been measured for NaHSO 4 , KHSO 4 and LiClO 4 ; 2) several compounds, including CH 3 SO 3 NH 4 , Mg(CH 3 SO 3 ) 2 , sodium malonate, sodium maleate and sodium tartrate, take up water vapor continuously as RH increases, and thus they do not exhibit distinctive deliquescence; and 3) the DRH values measured by different studies displayed large discrepancies for MgSO 4 , Ca(NO 3 ) 2 and Ca(ClO 4 ) 2 , precluding us from suggesting preferred DRH values for these compounds.

Table 1 Preferred deliquescence relative humidity (DRH) and efflorescence rela- tive humidity (ERH) at 298 K.
in Table1; for comparison, the preferred DRH values are provided for 99 compounds.There are also two other reasons why the preferred ERH values at 298 K are not given: (1) for some compounds, such as NaHSO 4 and Ca(NO 3 ) 2 , water vapor evaporates continuously DRH and ERH values are given in %. , and thus these compounds do not have well-defined ERH values; (2) compared to the DRH values, large discrepancies for ERH values measured by different studies occurred more frequently, and therefore we were unable to provide the preferred ERH values for these compounds.Below we discuss how the measured DRH ( Section 3.1 ) and ERH values ( Section 3.2 ) can be affected by environmental parameters and experimental conditions.