2.1.

Laser System

To overcome the protective skin barrier, the technique of laser microablation of the surface layers of skin was developed. The commercially available Palomar Lux2940 erbium laser (Palomar Medical Technologies Inc., Burlington, MA) was used as a light source. The laser parameters were as follows: the wavelength 2940 nm, the pulse energy 3.0 J, and the pulse duration 20 ms. The temporal profile of the laser pulse was modulated with three subpulses, with the subpulses’ duration of 5 ms. The laser emission beam was split into 169 microbeams using an array of microlenslets. Vertical microchannels (169) were created in the skin in the area with the dimensions of $6\times 6\mathrm{mm}$. The separation between the centers of the channels was $\sim 500\mu \text{m}$, diameter of their openings at skin surface was $\sim 100\mu \text{m}$, and the depth was $\sim 300\mu \text{m}$.

2.2.

Preparation of Particle Suspensions

We used the ${\mathrm{TiO}}_2$ nanopowder (634662-100G, Sigma-Aldrich Co., St. Louis, Missouri) consisting of a mixture of rutile and anatase ${\mathrm{TiO}}_2$ with the particle size $<100\mathrm{nm}$ and ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticles (LOMO PLC, Russia) with the particle size $\sim 27\mu \text{m}$.

Absorption spectra of the particles (the fraction of light absorbed by the particles) were measured with a commercially available spectrophotometer Lambda 950 (PerkinElmer Inc., Waltham, Massachusetts) with an integrating sphere in the spectral range 320 to 1000 nm. Measurement of sample absorption is possible using the center-mount option of the spectrophotometer, in which the sample is suspended in the middle of the integrating sphere.³⁴ The studied powders were put into a quartz cuvette with the dimensions $20\times 40\mathrm{mm}$ and the thickness 2 mm. The cuvette was suspended in the middle of the sphere at the angle of 15 deg in relation to incident beam to exclude the specular reflection.

To enhance the penetration depth of the particles into the skin and to suppress the photocatalytic action of photochemically active anatase form of ${\mathrm{TiO}}_2$, the particles were suspended in polyethylene glycol (PEG-300) having the molecular weight 300 (Sigma-Aldrich Co.). The pH of the chemical measured with a test-paper (Erba Lachema s.r.o., Czechia) was 6. The refractive index of the PEG-300 was measured with an Abbe refractometer (Atago DR-M2/1550, Japan) at several wavelengths (450, 589, 680, 1100, and 1550 nm) and interpolated. It was evaluated as 1.457 at 930 nm. Refractive indices of ${\mathrm{TiO}}_2$ and ${\mathrm{Al}}_2{\mathrm{O}}_3$ in the studied spectral range were 2.3 and 1.69, respectively.³⁵ The concentration of both nanoparticles and microparticles in the suspensions was $0.5\mathrm{mg}/\mathrm{mL}$. This concentration does not exceed the concentration of ${\mathrm{TiO}}_2$ and ${\mathrm{Al}}_2{\mathrm{O}}_3$ permitted in cosmetics. The content of ${\mathrm{TiO}}_2$ in personal care products and sunscreens is from 1% to $>10\%$ by weight.³⁰ ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticles were reported to be used in cosmetic products at concentrations up to 60%; and in skin care preparations, its concentration could be up to 25%.³³

The suspension was prepared using the CT-400A ultrasonic bath (CTBrand, Wan Luen Electronic Tools Co. Ltd., China) with the power of 35 W and the frequency of 43 to 45 kHz. The cuvette with the suspension was placed into the ultrasonic bath for 30 min for thorough mixing of the content both in the process of preparation and immediately before use.

2.3.

Nanoparticle Delivery

The ex vivo experiments were carried out with four samples of human skin extracted in the course of autopsy. The dimensions of the samples were approximately $60\times 20\mathrm{mm}$, and the average thickness was $3.5\pm 0.2\mathrm{mm}$. Each sample was divided into three sites. Both the first and the second sites were microporated. The third one remained intact. After microchannels were created, the ${\mathrm{TiO}}_2$ nanoparticle and ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticle suspensions were applied on the surface of the first site and the second site, respectively. The total exposure time was 2 h. During the procedure, the samples were placed in a Petri dish with a small volume of saline (aqueous solution of NaCl with the concentration $0.9\mathrm{mg}/\mathrm{mL}$) to prevent the change in the optical properties due to dehydration. Then, the suspensions were removed from the skin surface with a tampon wetted with saline. The remains of the suspensions were removed using the Multi-film medical sticky tape (Tesa, Germany) by two surface strips. The mean thickness of each strip of epidermis was about 0.5 μm.³⁶ The reflectance spectra from each site of the studied samples were measured. All experiments were carried out at room temperature ($\sim 20°\mathrm{C}$).

The in vivo study was carried out with three volunteers. The treated area was chosen on a forearm and was divided into three sites. Before commencing the experiment, the skin was disinfected with 70% ethanol. Both the first and the second sites were microporated. The third one remained intact. The ${\mathrm{TiO}}_2$ nanoparticle and ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticle suspensions were applied on the surface of the first site and the second site, respectively. The third one was also topically treated with ${\mathrm{TiO}}_2$ suspension. The solution of chlorhexidine was added to the suspension in the proportion 2:3 to prevent the infection of the target skin areas. After the fractional microablation, the particle application, and hand massage for 5 min, the treated sites were covered by polyethylene film and bandaged for 2 h. Then, the particles were removed from the skin surface by distilled water.

The observations were carried out over a month period: intact skin before the treatment, immediately after the FLMA, in 2 h, 24 h, 6 days, 14 days, and 30 days after the treatment.

Each observation included skin surface photography using the Nikon D80 camera (Nikon Inc., Japan) equipped with the Micro-Nikkor macro objective (Nikon Inc.), the measurements of the reflectance spectra from the studied areas of skin, and OCT scanning.

2.4.

Reflectance Measurements

The reflectance spectra of skin were measured using the USB4000-Vis-NIR multichannel spectrometer (Ocean Optics Inc., Dunedin, Florida) in the spectral range 380 to 1000 nm. The HL-2000 halogen lamp (Ocean Optics Inc., Dunedin, Florida) served as a source of light. The QR400-7-Vis/NIR fiber-optic probe (Ocean Optics Inc., Dunedin, Florida) consisting of seven fibers with the internal diameter 400 μm and the numerical aperture 0.2 was used in the measurements. The central fiber served for collecting the diffuse reflected radiation, while the surrounding six fibers were used for the illumination of the sample. The probe was placed at the distance of 2 mm from the skin surface and registered the signal averaged over the area of the radiation collection. The WS-1-SL reflectance standard (Ocean Optics Inc., Dunedin, Florida) was used to normalize the reflectance spectra.

2.5.

Optical Coherence Tomography

Visualization of microchannels filled with the suspension of the particles was implemented using a commercial OCT system (model OCP930SR, Thorlabs, USA)³⁷ working at the central wavelength $930\pm 5\mathrm{nm}$ with $100\pm 5\mathrm{nm}$ full width at half maximum (FWHM) spectrum, an optical power of 2 mW, a maximum image depth of 1.6 mm, and a length of scanned area 6 mm. The axial resolution (A-FWHM) determined by the spectral bandwidth of the light source was 6.2 μm in air. The lateral resolution (L-FWHM), entirely determined by the spatial focusing power of the objective, was 9.6 μm. In the medium, the parameters A-FWHM and L-FWHM increased.³⁸ If we suppose that the average refractive index of skin is about 1.42,³⁹ the axial and lateral resolutions can be evaluated as $\sim 4.5$ and $\sim 7\mu \text{m}$, respectively.

The optical probing depth of skin was evaluated from OCT A-scan as a level where intensity of OCT signal decreased by a factor of $e$.

2.6.

Histology Microscopic Study

Using the standard technique, the histological sections were prepared from the skin samples ex vivo. After fixation of the material with 10% solution of formalin, the samples were transferred through baths of progressively more concentrated ethanol to remove the water and then embedded in paraffin. The paraffined sections of 6 to 8 μm thickness were stained with haematoxylin and eosin.

The histological description of the preparations was carried out using the MC 100 XP microscope (Micros, Austria) operating in transmitted light mode with the magnification $200\times$. Photographs were taken using the Canon PC 1107 camera (Canon Inc., Japan).

3.1.

Histology Microscopic Study

Figure 1 presents microscopic images of histological sections of the skin sample ex vivo after FLMA with administering the ${\mathrm{TiO}}_2$ nanoparticle suspension into the skin. In the images in light background, the nanoparticle suspension appears as a dark inclusion (marked with arrows). Figure 1(a) shows an individual microchannel produced by FLMA. In this case, the suspension is observed as a thin layer along the walls of the channel. The depth of penetration of nanoparticles coincides with that of the channel. Figure 1(b) shows the area between the channels. As it can be seen, the unassisted topical application of the suspension has resulted in very superficial penetration of the compound, not extending beyond the boundaries of epidermis. The compound has mostly stayed on skin surface and filled the hollows in epidermal layer including openings of cutaneous sweat glands and follicles to the depth of less than 100 μm [Fig. 1(b)].

Fig. 1

Microscopic images of histological sections of the human skin samples ex vivo after the fractional laser microablation (FLMA) with administering the suspension into the skin (a) and unassisted topical application of ${\mathrm{TiO}}_2$ nanoparticle suspension (b). Bars correspond to 100 μm. Arrows indicate the nanoparticles localization.

3.2.

Ex vivo Experiments

Spectra of the light fraction absorbed in the studied nano- and microparticle powders in the spectral range of 320 to 1000 nm are presented in Fig. 2. ${\mathrm{TiO}}_2$ powder has a sharp absorption peak in the near UV spectral region and lacks any absorption in the visible and NIR spectral regions. ${\mathrm{Al}}_2{\mathrm{O}}_3$ has constant ($\sim 20\%$) absorption in the visible and NIR regions that increases slightly in the near UV.

Fig. 2

Absorption spectra of the ${\mathrm{TiO}}_2$ and ${\mathrm{Al}}_2{\mathrm{O}}_3$ powders.

Figure 3 shows reflectance spectra of skin ex vivo. The solid curve corresponds to the reflectance spectrum of the intact skin. The absorption bands of blood hemoglobin at the wavelengths 416 nm (Soret band), 543 nm, and 578 nm⁴⁰ are very pronounced. The superposition of absorption and scattering properties of the used particles with that of skin forms the reflectance spectra of skin doped with the particles. The dashed curve represents the spectrum of the skin site after the FLMA and administering the suspension of ${\mathrm{TiO}}_2$ nanoparticles. The depth of channels exceeded 300 μm; therefore, the suspension from the channels was not removed totally. The change in the skin spectrum shape indicates the presence of some amount of ${\mathrm{TiO}}_2$ nanoparticle suspension in the channels: the absorption band of ${\mathrm{TiO}}_2$ in the near UV spectral region (see Fig. 2) induced significant decrease of the skin reflectance in this spectral range. The increase in the intensity of radiation reflected from the skin was observed, which is related to the substantial increase of the scattering coefficient of the tissue, promoted by the nanoparticles located in the microchannels. The dotted line corresponds to the reflectance spectrum of the skin site after FLMA and administering ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticle suspension. In this case, the decrease of the reflectance in near UV spectral region is also observed in correspondence with absorption band of ${\mathrm{Al}}_2{\mathrm{O}}_3$ micropowder (see Fig. 2).

Fig. 3

Reflectance spectra of the ex vivo skin sample 2 h after FLMA and administering the ${\mathrm{TiO}}_2$ suspension (the site I) and ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension (the site II) into skin, and intact human skin (the site III).

3.3.

In vivo Experiments

Figure 4 represents a series of photographs of the human skin sites treated with FLMA and particle suspensions. Image dimension is $12\times 12\mathrm{mm}$. The contrast of all images was increased by 50% for better visual perception of the treated sites. For that we used a formula⁴¹: $I_{\text{out}}=(I_{\text{in}}-dI)·K+dI$, where $I_{\text{out}}$ is the new value of intensity, $I_{\text{in}}$ is the current value of intensity, $dI$ is the average value of intensity in an image, and $K$ is the contrast.

Fig. 4

Series of photographs of the human skin areas in vivo: site I was treated with FLMA and ${\mathrm{TiO}}_2$ suspension; site II was treated with FLMA and ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension; and site III was treated with ${\mathrm{TiO}}_2$ suspension. The images were obtained immediately after FLMA (a), 2 h (b); 24 h (c), 6 days (d), 14 days (e), and 30 days (f) after the treatments. Image size is $12\times 12\mathrm{mm}$.

The first column of the images corresponds to the site treated with FLMA and ${\mathrm{TiO}}_2$ suspension. The second one corresponds to the site treated with FLMA and ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension. The third one is the set of images of the site without FLMA. In this case, ${\mathrm{TiO}}_2$ suspension was applied on the skin surface topically. This site was included in the study to model the use of ${\mathrm{TiO}}_2$ as a component of sunscreens.

The images in the row (a) demonstrate the sites just after FLMA. The images Ia and IIa demonstrate well demarcated square treated areas resembling in appearance a lattice. Both erythema and slight edema in the areas of laser treatment were observed. In the image Ia, small lymph excretion took place. Blood excretion was not observed at either site.

In 2 h after the suspension application, the lightening of the skin was observed (see images Ib–IIIb). It was caused by the increase of skin reflection due to the particle suspension covering the surface. We also could see erythema and edema decreasing (see images Ib and IIb).

Twenty four hours later after the treatment, the complete healing of microchannels occurred. Treated sites were covered by a scab containing the particles. The openings of the channels filled up by the particles can be clearly seen (see images Ic and IIc). Some quantity of ${\mathrm{TiO}}_2$ nanoparticles could have stayed in the sulci of skin (image IIIc).

In 6 days after the treatment, peeling of the damaged surface layer of epidermis and the restoration of the epidermis integrity was observed (see images Id and IId). The scab on the skin surface remained. After 2 weeks, the microscopic channels were practically invisible; however, in the places of their location, slight redness was observed (see images Ie and IIe). In a month after the treatment, the pigmentation change was practically imperceptible (see images If and IIf). At this time point, no changes could be seen in the third site in comparison to baseline (images IIId–IIIf). It can be speculated that a total removing of the particles from the skin surface has occurred.

Figure 5 shows OCT images of the areas presented in Fig. 4. The depth and length of the scans correspond to 0.65 and 3 mm, respectively. In the images Ia and IIa, the sections of the ablated areas with microchannels are seen. The channels are marked with arrows. The image IIIa shows the OCT scan of the intact skin region. It is clearly seen that the skin microablation does not affect the optical depth of probing, which on average amounted to $\sim 320\mu \text{m}$, close to the value of this parameter for intact skin. In the image for intact skin, a papillary structure of epidermis on the surface of the skin and a boundary between epidermis and dermis are clearly seen.

Fig. 5

Series of optical coherence tomography (OCT) images of the human skin areas in vivo: site I was treated with FLMA and ${\mathrm{TiO}}_2$ suspension; site II was treated with FLMA and ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension; and site III was treated with ${\mathrm{TiO}}_2$ suspension. The images were obtained immediately after FLMA (a), 2 h (b); 24 h (c), 6 days (d), 14 days (e), and 30 days after the treatments. Bar corresponds to 1 mm.

In the images Ib and IIb, the arrows indicate that ablation channels are filled up with the particle suspensions at the 2 h timepoint. The apparent depth of the microchannel filled with the particles (it is marked in the image Ib with an arrow) can be evaluated as 300 to 310 μm. Since the average refractive index of dermis is 1.37,⁴² the actual geometrical depth of the channel is about 230 μm. The difference between channel depths ex vivo and in vivo can be explained by the following: the depth of penetration of nanoparticles can be noncoincident with that of the microscopic channels due to the healing of the channel walls in vital tissue. The image IIIb presents the skin covered with ${\mathrm{TiO}}_2$ suspension. The application of nanoparticle suspension induced some reduction of the optical probing depth (on average, up to 180 μm). The reduction of the optical probing depth may be explained by the shielding effect produced by particle layer, located in the channels, on the skin surface, and in the skin appendage openings. This effect was due to their high reflectivity. As a result, the OCT signal from the dermis tissue decreased significantly.

The next row of images was obtained 24 h after the removal of the suspension from the skin surface. It has become evident that the microchannels are filled up with the suspension in depth (see images Ic and IIc). In the image IIIc, we can also see natural sulci of skin filled up with ${\mathrm{TiO}}_2$ suspension. The localization of the particles is marked with arrows.

In the images Id and IId, which correspond to 6 days posttreatment, we can see the complete healing of microchannels. The peeling of the damaged surface layer of epidermis also begins at this point. On the skin surface, a scab containing particles (marked with arrows) forms a solid layer, which simulates a reflecting screen. Therefore, the optical probing depth decreases up to about 90 μm in the area of the microchannels and about 180 μm between them (see image Id) in the case of using the ${\mathrm{TiO}}_2$ suspension. When the ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension was applied (see image IId), the optical probing depth in the area of the microchannels was $\sim 100\mu \text{m}$, but in the area between them, the optical probing depth was equal with that of intact skin (see image IIId). It is possible that large ${\mathrm{Al}}_2{\mathrm{O}}_3$ particles were removed from the skin surface almost completely and remained only in the channels, whereas the ${\mathrm{TiO}}_2$ nanoparticles covered full treated surface.

Two weeks after the experiment, the peeling was completed (see images Ie and IIe). It can be suggested that bright sites in the OCT image Ie (marked with arrows) correspond to the closed microchannels still containing a quantity of nanoparticles. The image of intact skin (image IIIe) is identical to the initial one.

After the full restoration of the skin integrity (in a month), the particles, administered at the depth exceeding the epidermis thickness (100 to 150 μm), could stay in the dermis²⁵; however, their quantity was so little that they did not register in the OCT images. The optical probing depth for the images If, IIf, and IIIf was identical and coincided with that found before the experiment. However, sharp boundary between epidermis and dermis is absent in the images If and IIf unlike that in the image IIIf. In addition, blood vessels in the probing section of the skin are less visible. We hypothesize that the loss of the visualization contrast of epidermis and blood vessels in the OCT images can be related to the uniform distribution of the particles in the skin bulk.

Spectral measurements of skin (see Fig. 6) show differences in the reflectance of the studied skin sites related to the differences in fractions of light absorbed by the substances (see Fig. 2). We have chosen 24-h and 1-month time points for demonstration of changes in the reflectance skin spectra. Immediately after the laser treatment, both erythema and edema of the treated skin sites overshadowed the particle impact on the spectrum. In addition, the number of the particles on the skin surface was still too high to see the effects from in-depth inserted nanoparticles. Later on, erythema and edema decreased gradually on the both skin sites, but also the amount of the particles decreased due to the peeling of the upper epidermal layer. In a month, only particles remaining in deeper skin layers induced the changes in the spectra.

Fig. 6

Reflectance spectra of skin in vivo before (intact skin), 24 h posttreatment (a), and 30 days posttreatment (b) with laser microablation and administering the ${\mathrm{TiO}}_2$ suspension (site I), ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension (site II), and administering ${\mathrm{TiO}}_2$ suspension into intact skin (site III).

Figure 6(a) demonstrates the reflectance spectra of skin in vivo before and 24 h posttreatment. Figure 6(b) shows reflectance spectra of the same skin sites 1 month after the treatment. The solid curve corresponds to the spectrum of intact skin before any treatment. We can see characteristic bands of blood absorption and a band of water absorption.⁴⁰ Some shift of the maxima of the blood peaks in skin spectrum in vivo in comparison with that ex vivo (see Fig. 3) is caused by different content of oxy- and deoxyhemoglobin in blood at ex vivo and in vivo measurements. Dashed curves are the spectra of skin from the site I after the FLMA and administering the suspension of ${\mathrm{TiO}}_2$ nanoparticles. Dotted curves correspond to the spectra of skin from the site II after the FLMA and administering the suspension of ${\mathrm{Al}}_2{\mathrm{O}}_3$ microparticles. Dash-dot curves show the spectra of skin from the site III after an application of the suspension of ${\mathrm{TiO}}_2$ nanoparticles on skin surface.

In 24 h, we can see the changes of skin spectrum relatively to that of intact skin caused by both the microablation of epidermal layer and particle presence [see Fig. 6(a), sites I and II]. Erythema of the studied site (see Fig. 4, row c) is seen as spectral shift and magnitude increase of the band related to oxyhemoglobin absorption (417, 542, and 574 nm). Besides, on the treated site, residual edema was observed that is seen as decreasing reflectance intensity in the range of water absorption. ${\mathrm{TiO}}_2$ nanoparticles’ presence is revealed as the appearance of the signal decreasing in the spectral range shorter than 400 nm. Dotted curve corresponding to the spectrum of skin on the site II demonstrates additional lowering of signal in comparison with that on the site I. It can be related to absorption of ${\mathrm{Al}}_2{\mathrm{O}}_3$ suspension in the visible spectral range (see Fig. 2). Dash-dot curve shows the spectrum of skin from the site III. At this site, skin was not damaged; therefore, the shape of the reflectance spectrum in visible and NIR spectral ranges coincides with that of intact skin. ${\mathrm{TiO}}_2$ nanoparticles’ presence is revealed as the decreasing reflectance in the short-wavelength range as well as general increasing of the reflectance for other wavelengths due to increase of scattering from the skin surface.

At 1-month time point, decreasing of reflectance from the both sites I and II in comparison with that of intact skin suggests the presence of residual erythema [see Fig. 6(b)]. From the shape of the spectra in the range shorter than 400 nm, it is well seen that the particles are totally absent on the skin surface (dash-dot curve) and present in deeper skin layers (dashed and dotted curves). The spectral measurements correlate with OCT data on the in-depth localization of nanoparticles.