In recent years, photoinitiated polymerization has been widely used in the fields of UV-curable adhesives, UV-curable inks, UV-curable coatings, 3D printing, etc. The photopolymerization process is generally considered as a "green chemistry", which uses light as the driving force to form suitable initiating active species, such as free radicals, cations, etc., by absorbing photon energy and accompanying photochemical reactions, thus inducing the polymerization reaction. Therefore, it is very important to choose a suitable light-absorbing material, namely photoinitiator. The internal reason is that the absorption property (mainly wavelength and molar extinction coefficient) and reaction activity of the photoinitiator directly determine its initiation performance, and the external reason is that whether the absorption spectrum of the photoinitiator matches the emission spectrum of the light source directly affects the efficiency of the photoinitiator system. Next, we will explain the above concepts and their interrelationships one by one, hoping to enlighten the industry colleagues a little.

    Photoinitiator (PI) is the most important component of photocuring reaction and photocuring products. Figure 1 shows the molecular structure of some of the most commonly used photoinitiators in industrialization, including free-radical photoinitiators (divided into photocracking Type I, such as 1173, 184 and hydrogen extraction Type II, such as BP, ITX) and cationic photoinitiators (the last two iodonium salts and sulfonium salts).

    From the molecular structure of the above photoinitiator, we can see that the main light-absorbing group of the molecule is benzene ring, which determines that the maximum absorption wavelength of these photoinitiator molecules is basically in the UVB region (280-320nm, Figure 2). Take 907, which is widely used in the ink field, as shown in Figure 3. It can be seen from the figure that the wavelength of the maximum absorption peak of photoinitiator 907 in acetonitrile solvent is at 304nm (the absorption peak is related to the solvent and will move in the specific formula), and the maximum molar extinction coefficient ε 304nm~18000 L Mol-1 cm-1, it can be said that the absorption in this band is very strong. This traditional photoinitiator, represented by 907, is developed under the background that the excitation light source is a high pressure mercury lamp. The emission spectrum of the high pressure mercury lamp has a very strong peak at 302 nm (Fig. 2). Therefore, under the excitation of the mercury lamp, 907 photoinitiator has a very high customary efficiency and is a very efficient photoinitiator. So far, it has a large amount of use in fields such as photocuring ink. However, when entering the era of UV-LED, we found that the use of many traditional photoinitiators such as 907 has been greatly restricted. The reason is that the emission wavelength of LED is very different from that of mercury lamp. When we observe the absorption of 907 in the UV-LED band, we found that its absorbance is very small, for example ε 365nm~100 L Mol-1 cm-1。

    We know that the emission spectra of UVLED light sources with relatively mature industrial applications are mainly located in the UVA band and the longer wavelength region, that is, 365-405 nm. It is obvious that most of the traditional photoinitiators are seriously mismatched with the emission spectra of LED light sources. In fact, in addition to TPO, 819, ITX and other initiators have good absorption in this region, the absorption of other initiators in the UVA band is very small. This can be well explained in Figure 3 (right).

    So can we make LED light source that matches the absorption wavelength of traditional photoinitiator? Theoretically, any wavelength of LED light source can be made. As shown in Figure 4, by changing the proportion of semiconductor materials, LED chips with different emission wavelengths can be made.

    Apart from the most difficult chip manufacturing, the UVLED light source alone involves the selection of chip specifications, chip packaging, system optical design, power system selection, cooling system scheme design, and so on. If any link is not done well, it will affect the stability and reliability of the entire system. So although theoretically any wavelength of LED light source can be made, it is very limited that it can be truly industrialized.

    Due to the narrow emission spectrum of LED, the half-peak width (60% of the luminous power) is generally within 10 nm. Taking 365nm and 395nm LEDs as examples, we found that the overlap of the absorption spectrum with 907 is very small, which leads to the weak overlap integration of the absorption spectrum and the emission spectrum of the photoinitiator, so most of the energy of the light source is wasted and cannot be used for the luminescence chemical reaction. However, the actual photopolymerization effect proves that these photoinitiators can effectively initiate polymerization when LED is used as the light source. Therefore, we must consider other factors in the formula. Internal and external causes need to be considered comprehensively.

    This is because UV curing is a complex system that interacts from the formula to the light source, and the molar extinction coefficient is only a factor that affects the absorbance of the formula. Factors such as light intensity, film thickness and initiator concentration must also be considered. After a formula is coated into a film, its absorbance (A) conforms to Langer-Beer law:

    Among them, ε Is the molar extinction coefficient related to wavelength, as shown in Figure 1, which is different at different wavelengths. Generally speaking, when ε If the value of is greater than 100L Mol-1cm-1, the photoinitiator is valuable, for example, the molar extinction coefficient of 907 at 365nm is in this range. A special kind of visible light initiator camphor quinone (CQ), which has a value of only 29 L Mol-1 in the visible light region, has very good application effect in fields such as light curing. The reason lies in the following two parameters. Here c is the molar concentration. In the actual formula, the mass concentration is generally used. The mass content of the general photoinitiator is 2%~6%, generally not more than 8%~10%. Taking the varnish formula as an example, the density of the diluent and resin is assumed to be 1 g/mL, and c is about 0.067~0.33 mol/L, which is a relatively high value. In the acetonitrile solvent in Figure 2, c is about 10-4 mol/L, so the concentration in the formula is 100-1000 times. Therefore, The absorption of UVA will increase significantly; In the formula, l is the thickness of the film, in cm. The thickness of the coating applied in practice is micron, generally 5~50 μ M range. So the absorbance A of the formula can be calculated by the formula, about 0.02~2.0. The proportion of light absorbed by the formula film can be calculated by A value:

    According to formula (2), substituting A=0.02~2.0, we can calculate that the proportion of light absorbed by the film is about 5%~99% in the absorbance range. When the formula coating film is only 5 microns thick, even if the concentration of the initiator increases to 10wt%, only 5% of the light will be used, and 95% will not work (it can be reused through the light path design reflection of LED equipment, which is a practical problem, not included in this theoretical calculation process). When the initiator concentration is too high or the film is thick, the upper formula can absorb 99% of the photons, and the lower formula has no light irradiation, and the curing effect is poor, such as adhesion. Therefore, a good formula must comprehensively consider these three factors. However, in practical applications, the content of photoinitiator is often increased to overcome oxygen polymerization inhibition, light source mismatch and so on. On the one hand, the cost is increased, and on the other hand, the quality is reduced, such as the residual of initiator, the migration of small molecules, odor, etc.

    In order to overcome these weaknesses, the internal cause can be improved by designing a new photoinitiator system or improving, improving and expanding the photoactivity, spectral response range and other properties of existing initiators, which is also a research hotspot in academia and industry. The main trend of the research and development of new photoinitiators is to have good light absorption properties in the long wavelength region (UVA to visible light), so that the compatibility with LED will be significantly improved. Figure 5 (a) shows the absorption spectrum of a new pyrazoline sulfonium salt. The maximum absorption wavelength of the molecule shifts to 342 nm, and the molar extinction coefficient at this wavelength is 15400 L Mol-1 cm-1. It can be inferred that the absorption at 365nm LED will be significantly enhanced, as shown in Figure 5b.