**HRTEM Study on Resistive Switching ZrO2 Thin Films and Their Micro-Fabricated Thin Films**

Ying Li, Gaoyang Zhao, Zhibo Kou, Long Jin and Yajing Wang

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/60973

#### **Abstract**

For the next-generation nonvolatile memory material, the most promising candi‐ date is resistive random access memory (RRAM) which is nonvolatile memory with high density, high speed, and low power consumption. Resistive switching (RS) be‐ havior had been reported in various films including transition metal oxides, perov‐ skite, and chalcogenide. For further application, it is still a challenge to fabricate nanostructures of RS material. Micro-fabrication method involves traditional lithog‐ raphy, chemical etching, electron beam direct writing, nano-imprint, and so on. However, the procedure and the cost of these methods are relatively complex and high for semiconductors process. In this chapter, we demonstrate a method for fab‐ ricating sub-micro ZrO2 lattice by using sol-gel method combined with laser inter‐ ference lithography and micro-analysis with high-resolution transmission electron microscopy (HRTEM).

Researchers used all kinds of techniques to investigate the mechanism such as con‐ ductive atomic force microscopy, HRTEM, and scanning electrical microscopy. De‐ spite the extensive research, much of the underlying mechanism is still unclear and controversial. This task can be accomplished only with advanced measurement and technique such as in HRTEM, local conductive atomic force microscopy, and so on. HRTEM sample preparation method for array dots is also discussed. In our re‐ search, the bipolar RS behavior can be observed successfully in this structure. And HRTEM observation was used to study the interface between the layers. RRAM unit consists of a conductive atomic force microscope tip as an anode, a ZrO2 lattice dot as RS material, and a copper electrode as a cathode.

**Keywords:** Resistive switching, HRTEM, Sample preparation

© 2015 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **1. Introduction**

The development of information technology has brought rapid changes to people's life. Re‐ sistive random access memory (RRAM), because of its fast response, high stability, low power consumption, with the traditional advantages of the semiconductor technology, has good compatibility, and according to scientists, is expected to become one of the next gener‐ ations of memory [1-5]. In this chapter, by using shadow lithography and laser interference lithography, two kinds of methods to micro-fabricate the RS oxide film, it is suggested that the oxide film is suitable for micromachining process than using atomic force microscopy (AFM) in situ electrical measurement technique for fine-patterned films after the oxide film was studied by means of in situ micro area electrical performance. In this chapter, we used sol-gel method and chemical modification method for the preparation of a variety of ZrO2 sol, and optimizing the good comprehensive properties of oxide RS material, through the analysis of the HRTEM studies of the oxide film [6-10].

### **2. Experiment**

#### **2.1. ZrO2 thin films preparation**

Photosensitive ZrO2 films were fabricated with the metal alkoxide zirconium tetran-butox‐ ide [Zr(OC4H9)4], benzoylacetone (BzAcH), and ethanol (EtOH) in the molar ratio of Zr:BzAcH:EtOH at 1:1:40. After being mixed and refluxed for 2 to 3 hours, the photosensi‐ tive ZrO2 solution was obtained. Then using dip-coating process, ZrO2 organic films were formed on copper substrate. The copper substrates were fabricated by magnetic sputtering. on copper substrate. The copper substrates were fabricated by magnetic sputtering.

#### **2.2. ZrO2 lattice preparation**

2.2 ZrO2 lattice preparation

dissociated by UV irradiation.

ZrO2 lattice was fabricated by double exposures of two-beam laser interference through 90° rotation of the sample in its own plane between twice exposures with 350 nm Krypton ion laser. After laser exposure, the exposed part of ZrO2 film is not soluble in leaching solution while the ZrO2 lattice was fabricated by double exposures of two-beam laser interference through 90° rotation of the sample in its own plane between twice exposures with 350 nm Krypton ion laser. After laser exposure, the exposed part of ZrO2 film is not soluble in leaching solution while the un-exposed part is still soluble. Therefore, ZrO2 lattice can be obtained [11-14].

un-exposed part is still soluble. Therefore, ZrO2 lattice can be obtained [11-14]. The UV–visible absorption spectra were measured and these spectra of the The UV–visible absorption spectra were measured and these spectra of the ZrO2 films were dip-coated on quartz glass substrate. It was known that an absorption peak of BzAcH at 308 nm was due to the π-to-π\* transition. These peaks shift to 338 nm in the spectra of the ZrO2 gel films. This red shift indicates the chelation between Zr4+ and BzAcH, and the reaction re‐

ZrO2 films were dip-coated on quartz glass substrate. It was known that an absorption peak of BzAcH at 308 nm was due to the π-to-π\* transition. These peaks shift to 338 nm in the spectra of the ZrO2 gel films. This red shift indicates the chelation between Zr4+ and BzAcH, and the reaction results in the partial replacement of the alkoxy group by β-diketone ligand [see Eq. (1)]. Fig. 1 also shows the AFM image of ZrO2 films irradiated by a UV Krypton ion laser. The intensity of absorption peaks at around 338 nm decreases step by step with increasing irradiation time from 0 second to 400 seconds. These decreasing spectra showed that the chelate rings were sults in the partial replacement of the alkoxy group by β-diketone ligand [see Eq. (1)]. Fig. 1 also shows the AFM image of ZrO2 films irradiated by a UV Krypton ion laser. The intensity of absorption peaks at around 338 nm decreases step by step with increasing irradiation time from 0 second to 400 seconds. These decreasing spectra showed that the chelate rings were dissociated by UV irradiation.

**Figure 1.** Observation of ZrO2 lattice dot by AFM measurement systems.

### **2.3. I-V measurement with ZrO2 thin films**

**1. Introduction**

**2. Experiment**

sputtering.

[11-14].

**2.1. ZrO2 thin films preparation**

The development of information technology has brought rapid changes to people's life. Re‐ sistive random access memory (RRAM), because of its fast response, high stability, low power consumption, with the traditional advantages of the semiconductor technology, has good compatibility, and according to scientists, is expected to become one of the next gener‐ ations of memory [1-5]. In this chapter, by using shadow lithography and laser interference lithography, two kinds of methods to micro-fabricate the RS oxide film, it is suggested that the oxide film is suitable for micromachining process than using atomic force microscopy (AFM) in situ electrical measurement technique for fine-patterned films after the oxide film was studied by means of in situ micro area electrical performance. In this chapter, we used sol-gel method and chemical modification method for the preparation of a variety of ZrO2 sol, and optimizing the good comprehensive properties of oxide RS material, through the

Photosensitive ZrO2 films were fabricated with the metal alkoxide zirconium tetran-butox‐ ide [Zr(OC4H9)4], benzoylacetone (BzAcH), and ethanol (EtOH) in the molar ratio of Zr:BzAcH:EtOH at 1:1:40. After being mixed and refluxed for 2 to 3 hours, the photosensi‐ tive ZrO2 solution was obtained. Then using dip-coating process, ZrO2 organic films were formed on copper substrate. The copper substrates were fabricated by magnetic sputtering.

on copper substrate. The copper substrates were fabricated by magnetic

ZrO2 lattice was fabricated by double exposures of two-beam laser interference through 90° rotation of the sample in its own plane between twice exposures with 350 nm Krypton ion laser. After laser exposure, the exposed part of ZrO2 film is not soluble in leaching solution while the un-exposed part is still soluble. Therefore, ZrO2 lattice can be obtained

ZrO2 lattice was fabricated by double exposures of two-beam laser interference through 90° rotation of the sample in its own plane between twice exposures with 350 nm Krypton ion laser. After laser exposure, the exposed part of ZrO2 film is not soluble in leaching solution while the un-exposed part is still soluble. Therefore, ZrO2 lattice can be obtained [11-14].

The UV–visible absorption spectra were measured and these spectra of the ZrO2 films were dip-coated on quartz glass substrate. It was known that an absorption peak of BzAcH at 308 nm was due to the π-to-π\* transition. These peaks shift to 338 nm in the spectra of the ZrO2 gel films. This red shift indicates the chelation between Zr4+ and BzAcH, and the reaction results in the partial replacement of the alkoxy group by β-diketone ligand [see Eq. (1)]. Fig. 1 also shows the AFM image of ZrO2 films irradiated by a UV Krypton ion laser. The intensity of absorption peaks at around 338 nm decreases step by step with increasing irradiation time from 0 second to 400 seconds. These decreasing spectra showed that the chelate rings were

The UV–visible absorption spectra were measured and these spectra of the ZrO2 films were dip-coated on quartz glass substrate. It was known that an absorption peak of BzAcH at 308 nm was due to the π-to-π\* transition. These peaks shift to 338 nm in the spectra of the ZrO2 gel films. This red shift indicates the chelation between Zr4+ and BzAcH, and the reaction re‐

(C4H9O)

O

C

CH

(C + C4H9OH 4H9O)

C6H5

CH3

Eq.

C

O

analysis of the HRTEM studies of the oxide film [6-10].

180 The Transmission Electron Microscope – Theory and Applications

3Zr (C4H9O)4Zr + <sup>H</sup>

CH3

CH2

C6H5

C

C

O

O

2.2 ZrO2 lattice preparation

**2.2. ZrO2 lattice preparation**

dissociated by UV irradiation.

Before the local conductive AFM (LC-AFM) measurement of the ZrO2 lattice dots, a tradi‐ tional I-V probe station measurement was carried out with the normal ZrO2 thin film which is made of the photosensitive ZrO2 solution. This RS structure is formed with platinum (top electrode), ZrO2 photosensitive film (RS layer), and copper (bottom electrode). The platinum top electrode is made by sputtering with shadow mask. The pattern of the shadow mask is periodic lattice dot with a diameter of 2 mm.

### **2.4. I-V measurement with ZrO2 lattice**

This LC-AFM measurement system consists an atomic force microscope (AFM, Seiko Instru‐ ments Inc, SPI3800N/SPA-400) with a resolution of 0.1 nm in x–y directions and 0.01 nm in the z direction and a Keithly 2400 I-V source meter (Keithly, USA.). The silicon tips (PPP-CONT, Nanosensors) were coated by sputtering deposition with a 25 nm thick double layer of chromium and platinum/Iridium (Pt/Ir) with an initial 20 nm radius. In our LC-AFM measurement study, RRAM unit consists of a conductive atomic force microscope tip as an anode, a ZrO2 lattice dot as RS material, and a copper electrode as a cathode. In order to identify the RS properties of ZrO2 array dot, the Pt/Ir conductive atomic force microscope tip is tightly positioned at the center of one of the dots. During the LC-AFM measurement, Pt/Ir conductive probe and the base of the LC-AFM are connected to Keithly 2400 meter to apply the excitation signal and record voltage and current response of the RS array dots. When the voltages from Keithly 2400 meter are applied to the Pt/Ir conductive AFM tip and the bot‐ tom electrode of the sample, the current flow occurs through the cantilever and goes back to the Keithly 2400 meter. Then a local I-V curve can be obtained.

In order to solve the influence from huge contact resistance between the Pt/Ir conductive atomic force microscope tip and a single ZrO2 array dot, a platinum layer was spread on the whole ZrO2 lattice by sputtering. In this way, the contact can be improved.

### **2.5. HRTEM sample preparation**

The HRTEM observation was carried out by HRTEM (JEOL 3010, Japan). The diagram of sample preparation process is shown in Fig. 2. In Fig. 2(a), we cut some tiny part from the platinum-covered sample which can achieve I-V curves by LC-AFM systems for HRTEM ob‐ servations.

Two tiny parts were pasted by AB glue face-to-face. AB glue should be used as less as possi‐ ble because the more AB glue used the worse ion milling would be. And this unit was mil‐ led by hand in x direction as smooth as possible. When the thickness in z direction decreases down to less than 1 μm, this unit can be pasted to nickel (Ni) grid. This will be ready for Argon ions milling. A cross-section sample of Pt/ZrO2 lattice dot/Cu unit was milled by the low-angle Argon milling machine (Gatan, USA). The milling angle was less than 8°. Micro‐ structural analysis was done on the basis of HRTEM observations. It is well known that the preparation of good specimens for high-resolution transmission microscopy is a very impor‐ tant process. The milling angle of Argon ions was less than 8° for the reasons of the damage from Argon ions. The transparent area to electrons can be widened if the Argon ions milling angle was small. However, it cannot mill if the milling angle was less than 3°. The ideal mill‐ ing angle of Argon ions was between 4° and 8°.

HRTEM Study on Resistive Switching ZrO2 Thin Films and Their Micro-Fabricated Thin Films http://dx.doi.org/10.5772/60973 183

**Figure 2.** Illustration of HRTEM samples preparation process.

### **3. Discussion and results**

is made of the photosensitive ZrO2 solution. This RS structure is formed with platinum (top electrode), ZrO2 photosensitive film (RS layer), and copper (bottom electrode). The platinum top electrode is made by sputtering with shadow mask. The pattern of the shadow mask is

This LC-AFM measurement system consists an atomic force microscope (AFM, Seiko Instru‐ ments Inc, SPI3800N/SPA-400) with a resolution of 0.1 nm in x–y directions and 0.01 nm in the z direction and a Keithly 2400 I-V source meter (Keithly, USA.). The silicon tips (PPP-CONT, Nanosensors) were coated by sputtering deposition with a 25 nm thick double layer of chromium and platinum/Iridium (Pt/Ir) with an initial 20 nm radius. In our LC-AFM measurement study, RRAM unit consists of a conductive atomic force microscope tip as an anode, a ZrO2 lattice dot as RS material, and a copper electrode as a cathode. In order to identify the RS properties of ZrO2 array dot, the Pt/Ir conductive atomic force microscope tip is tightly positioned at the center of one of the dots. During the LC-AFM measurement, Pt/Ir conductive probe and the base of the LC-AFM are connected to Keithly 2400 meter to apply the excitation signal and record voltage and current response of the RS array dots. When the voltages from Keithly 2400 meter are applied to the Pt/Ir conductive AFM tip and the bot‐ tom electrode of the sample, the current flow occurs through the cantilever and goes back to

In order to solve the influence from huge contact resistance between the Pt/Ir conductive atomic force microscope tip and a single ZrO2 array dot, a platinum layer was spread on the

The HRTEM observation was carried out by HRTEM (JEOL 3010, Japan). The diagram of sample preparation process is shown in Fig. 2. In Fig. 2(a), we cut some tiny part from the platinum-covered sample which can achieve I-V curves by LC-AFM systems for HRTEM ob‐

Two tiny parts were pasted by AB glue face-to-face. AB glue should be used as less as possi‐ ble because the more AB glue used the worse ion milling would be. And this unit was mil‐ led by hand in x direction as smooth as possible. When the thickness in z direction decreases down to less than 1 μm, this unit can be pasted to nickel (Ni) grid. This will be ready for Argon ions milling. A cross-section sample of Pt/ZrO2 lattice dot/Cu unit was milled by the low-angle Argon milling machine (Gatan, USA). The milling angle was less than 8°. Micro‐ structural analysis was done on the basis of HRTEM observations. It is well known that the preparation of good specimens for high-resolution transmission microscopy is a very impor‐ tant process. The milling angle of Argon ions was less than 8° for the reasons of the damage from Argon ions. The transparent area to electrons can be widened if the Argon ions milling angle was small. However, it cannot mill if the milling angle was less than 3°. The ideal mill‐

periodic lattice dot with a diameter of 2 mm.

182 The Transmission Electron Microscope – Theory and Applications

the Keithly 2400 meter. Then a local I-V curve can be obtained.

whole ZrO2 lattice by sputtering. In this way, the contact can be improved.

**2.4. I-V measurement with ZrO2 lattice**

**2.5. HRTEM sample preparation**

ing angle of Argon ions was between 4° and 8°.

servations.

#### **3.1. Normal probe station I-V curves**

For the electrical measurements, probes contact top electrode (Cu) and bottom electrode SnO2:Sb (ATO), respectively. The bottom electrode was grounded. In Fig. 3, it shows the I-V characteristics of the Cu/ZrO2/ATO structure after fifty times of I-V measurements cycles. The current compliance (CC) is applied to 10 mA during I-V sweep. When the voltage goes to 0.5 V, the current increases rapidly which means the OFF state switches to the ON state. During the I-V sweep at negative side, when the voltage goes to -1.4 V, the current decreases rapidly which means the ON state switches to the OFF state. This behavior can be repeated more than fifty times.

This behavior can be repeated more than fifty times.

0.5 V, the current increases rapidly which means the OFF state switches to the ON state. During the I-V sweep at negative

side, when the voltage goes to -1.4 V, the current decreases rapidly which means the ON state switches to the OFF state.

decreased when the RESET voltage value decreased. For I-V curve in Fig. 3, voltage sweep started from 0 V to the -2 V and

During the LC-AFM measurement, the parameters of I Gain and P Gain should be adjusted. The value of P Gain should

be set as one-third of I Gain value. The Pt/Ir conductive coating could be switched off when the value of Force Reference is

too small. This value should be set from -1 nm to -4 nm. The sweep voltage can be set from -25 V to 25 V. The height of

array dot A (ΔZ) is about 61 nm and dot B (ΔZ) is about 58 nm. The diameter of dot A is about 1.2 μm and that of dot B is

about 1.1 μm. Local I-V curves are measured while the LC-AFM image was scanning. The CC is applied to 100 μA during

I-V sweep. When the voltage goes to 18 V, the current increases rapidly which means the OFF state switches to the ON

state. During the I-V sweep at negative side, when the voltage goes to -5.0 V, the current decreases rapidly which means

the ON state switches to the OFF state. In order to make sure that the array dot is not shifted during the LC-AFM

measurement, an AFM image scan is very necessary after the local I-V measurement is finished.

**Figure 3.** I-V curve of the Cu/ZrO2/ATO structure after fifty cycles.

Figure 3. I-V curve of the Cu/ZrO2/ATO structure after fifty cycles. Bipolar RS characterization was stable for ZrO2 thin film in the structure unit. Normally, the voltage sweeps start from 0 V and go to the positive side (1 V) and goes back to 0 V. Resistance switches from high resistance state (RH) to low resistance state (RL) and it is called SET process. And then the voltage sweeps from 0 V to the negative side (-2 V) and goes back to 0 Bipolar RS characterization was stable for ZrO2 thin film in the structure unit. Normally, the voltage sweeps start from 0 V and go to the positive side (1 V) and goes back to 0 V. Resist‐ ance switches from high resistance state (RH) to low resistance state (RL) and it is called SET process. And then the voltage sweeps from 0 V to the negative side (-2 V) and goes back to 0 V. Resistance switches from RL to RH and it is called RESET process. When the voltage swept at -0.5 V, RL was 70 Ω and RH was 100 KΩ. This behavior can be repeated hundreds of times. The CC is applied with 10 mA. The highest ratio of RH:RL can reach to 104 .

V. Resistance switches from RL to RH and it is called RESET process. When the voltage swept at -0.5 V, RL was 70 Ω and RH was 100 KΩ. This behavior can be repeated hundreds of times. The CC is applied with 10 mA. The highest ratio of RH:RL can reach to 104. During the RESET operation, the interesting thing we found is that RH varied when we changed the RESET voltage. RH decreased when the RESET voltage value decreased. For I-V curve in Fig. 3, voltage sweep started from 0 V to the -2 V and then from -2 V to 0 V during the RESET process. When the voltage swept at -0.5 V, RL was 100 Ω and RH was 250 KΩ. This behavior can be repeated more than several times. The CC is applied with 10 mA. The highest ratio of RH:RL can reach to 104 .

#### During the RESET operation, the interesting thing we found is that RH varied when we changed the RESET voltage. RH **3.2. Local I-V curves by LC-AFM measurement systems**

then from -2 V to 0 V during the RESET process. When the voltage swept at -0.5 V, RL was 100 Ω and RH was 250 KΩ. This behavior can be repeated more than several times. The CC is applied with 10 mA. The highest ratio of RH:RL can reach to 104. **3.2. Local I-V curves by LC-AFM measurement systems** During the LC-AFM measurement, the parameters of I Gain and P Gain should be adjusted. The value of P Gain should be set as one-third of I Gain value. The Pt/Ir conductive coating could be switched off when the value of Force Reference is too small. This value should be set from -1 nm to -4 nm. The sweep voltage can be set from -25 V to 25 V. The height of array dot A (ΔZ) is about 61 nm and dot B (ΔZ) is about 58 nm. The diameter of dot A is about 1.2 μm and that of dot B is about 1.1 μm. Local I-V curves are measured while the LC-AFM im‐ age was scanning. The CC is applied to 100 μA during I-V sweep. When the voltage goes to

18 V, the current increases rapidly which means the OFF state switches to the ON state. Dur‐ ing the I-V sweep at negative side, when the voltage goes to -5.0 V, the current decreases rapidly which means the ON state switches to the OFF state. In order to make sure that the array dot is not shifted during the LC-AFM measurement, an AFM image scan is very neces‐ sary after the local I-V measurement is finished. A platinum film of 30 nm thickness was sputtered on ZrO2 lattice dot. Fig. 4 is a local I-V curve which is also measured

A platinum film of 30 nm thickness was sputtered on ZrO2 lattice dot. Fig. 4 is a local I-V curve which is also measured while the LC-AFM image was scanning. The CC is applied to 100 μA during I-V sweep. When the voltage goes to 0.8 V, the current increases rapidly which means the OFF state switches to the ON state. During the I-V sweep at negative side, when the voltage goes to -2.2 V, the current decreases rapidly which means the ON state switches to the OFF state. while the LC-AFM image was scanning. The CC is applied to 100 μA during I-V sweep. When the voltage goes to 0.8 V, the current increases rapidly which means the OFF state switches to the ON state. During the I-V sweep at negative side, when the voltage goes to -2.2 V, the current decreases rapidly which means the ON state switches to the OFF state.

**Figure 4.** A local I-V curve which is measured while the LC-AFM image was scanning.

copper electrode to its high resistance state [15-19].

Figure 5. HRTEM image of Cu/ZrO2/ATO structure.

#### During the RESET operation, the interesting thing we found is that RH varied when we changed the RESET voltage. RH Figure 4. A local I-V curve which is measured while the LC-AFM image was scanning. **3.3. HRTEM observations**

0.5 V, the current increases rapidly which means the OFF state switches to the ON state. During the I-V sweep at negative

side, when the voltage goes to -1.4 V, the current decreases rapidly which means the ON state switches to the OFF state.

During the LC-AFM measurement, the parameters of I Gain and P Gain should be adjusted. The value of P Gain should

.

be set as one-third of I Gain value. The Pt/Ir conductive coating could be switched off when the value of Force Reference is

too small. This value should be set from -1 nm to -4 nm. The sweep voltage can be set from -25 V to 25 V. The height of

array dot A (ΔZ) is about 61 nm and dot B (ΔZ) is about 58 nm. The diameter of dot A is about 1.2 μm and that of dot B is

about 1.1 μm. Local I-V curves are measured while the LC-AFM image was scanning. The CC is applied to 100 μA during

I-V sweep. When the voltage goes to 18 V, the current increases rapidly which means the OFF state switches to the ON

state. During the I-V sweep at negative side, when the voltage goes to -5.0 V, the current decreases rapidly which means

the ON state switches to the OFF state. In order to make sure that the array dot is not shifted during the LC-AFM

measurement, an AFM image scan is very necessary after the local I-V measurement is finished.

This behavior can be repeated more than fifty times.

Figure 3. I-V curve of the Cu/ZrO2/ATO structure after fifty cycles.

The CC is applied with 10 mA. The highest ratio of RH:RL can reach to 104

.

**3.2. Local I-V curves by LC-AFM measurement systems**

Bipolar RS characterization was stable for ZrO2 thin film in the structure unit. Normally, the voltage sweeps start from 0 V and go to the positive side (1 V) and goes back to 0 V. Resist‐ ance switches from high resistance state (RH) to low resistance state (RL) and it is called SET process. And then the voltage sweeps from 0 V to the negative side (-2 V) and goes back to 0 V. Resistance switches from RL to RH and it is called RESET process. When the voltage swept at -0.5 V, RL was 70 Ω and RH was 100 KΩ. This behavior can be repeated hundreds of times.

During the RESET operation, the interesting thing we found is that RH varied when we changed the RESET voltage. RH decreased when the RESET voltage value decreased. For I-V curve in Fig. 3, voltage sweep started from 0 V to the -2 V and then from -2 V to 0 V during the RESET process. When the voltage swept at -0.5 V, RL was 100 Ω and RH was 250 KΩ. This behavior can be repeated more than several times. The CC is applied with 10 mA. The

During the LC-AFM measurement, the parameters of I Gain and P Gain should be adjusted. The value of P Gain should be set as one-third of I Gain value. The Pt/Ir conductive coating could be switched off when the value of Force Reference is too small. This value should be set from -1 nm to -4 nm. The sweep voltage can be set from -25 V to 25 V. The height of array dot A (ΔZ) is about 61 nm and dot B (ΔZ) is about 58 nm. The diameter of dot A is about 1.2 μm and that of dot B is about 1.1 μm. Local I-V curves are measured while the LC-AFM im‐ age was scanning. The CC is applied to 100 μA during I-V sweep. When the voltage goes to

**3.2. Local I-V curves by LC-AFM measurement systems**

can reach to 104.

highest ratio of RH:RL can reach to 104

**Figure 3.** I-V curve of the Cu/ZrO2/ATO structure after fifty cycles.

184 The Transmission Electron Microscope – Theory and Applications

104.

decreased when the RESET voltage value decreased. For I-V curve in Fig. 3, voltage sweep started from 0 V to the -2 V and then from -2 V to 0 V during the RESET process. When the voltage swept at -0.5 V, RL was 100 Ω and RH was 250 KΩ. This behavior can be repeated more than several times. The CC is applied with 10 mA. The highest ratio of RH:RL can reach to **3.3. HRTEM observations** We took one unit of samples for TEM observations in which the ZrO2 thin film in the structure was annealed at 300°C. Microstructural analysis has been done on the basis of TEM observations. Fig. 5 shows the interface image of Cu/ZrO2/ATO structure in which the ZrO2 thin film was annealed at 300°C. The bottom electrode (BE) is ATO film and the thickness is about 400 nm. The ZrO2 work layer is about 200 nm thick and it is amorphous according to the selected area We took one unit of samples for TEM observations in which the ZrO2 thin film in the struc‐ ture was annealed at 300°C. Microstructural analysis has been done on the basis of TEM ob‐ servations. Fig. 5 shows the interface image of Cu/ZrO2/ATO structure in which the ZrO2 thin film was annealed at 300°C. The bottom electrode (BE) is ATO film and the thickness is about 400 nm. The ZrO2 work layer is about 200 nm thick and it is amorphous according to the selected area diffraction pattern. In Fig. 5, it is shown that copper ions are penetrated into the ZrO2 layer. It can be proved by the Fast Fourier Transform patterns. According to the filament formation mechanism, resistance reduction in the devices is due to the existing

> diffraction pattern. In Fig. 5, it is shown that copper ions are penetrated into the ZrO2 layer. It can be proved by the Fast Fourier Transform patterns. According to the filament formation mechanism, resistance reduction in the devices is due to the existing copper to form conducting copper-rich pathways. An opposite bias takes the existing copper back to the

> Fig. 6(a) shows the interface image of Pt/ZrO2 lattice dot/Cu sandwiched structure. Two lattice dots were seen in this figure, which are labeled as A and B. These two dots were not similar in size. This phenomenon was also seen in the rest of the HRTEM sample observations. The diameter of dot A is around 800 nm while that of dot B is around 700 nm. This is not agreeing with the LC-AFM image results. Through LC-AFM image observation, it is shown that the ZrO2 array dots are even and in the same diameter range of about 1 μm. This is because when the HRTEM sample was milling manually

copper to form conducting copper-rich pathways. An opposite bias takes the existing copper back to the copper electrode to its high resistance state [15-19]. conducting copper-rich pathways. An opposite bias takes the existing copper back to the copper electrode to its high resistance state [15-19].

Fig. 5. HRTEM image of Cu/ZrO2/ATO structure. **Figure 5.** HRTEM image of Cu/ZrO2/ATO structure.

Fig. 6(a) shows the interface image of Pt/ZrO2 lattice dot/Cu sandwiched structure. Two lattice dots were seen in this figure, which are labeled as A and B. These two dots were not similar in size. This phenomenon was also seen in the rest of the HRTEM sample observations. The diameter of dot A is around 800 nm while that of dot B is around 700 nm. This is not agreeing with the LC-AFM image results. Through LC-AFM image observation, it is shown that the ZrO2 array dots are even and in the same diameter range of about 1 µm. This is because when the HRTEM sample was milling manually in the first stage, it was difficult to keep the milling plane vertical Fig. 6(a) shows the interface image of Pt/ZrO2 lattice dot/Cu sandwiched structure. Two lat‐ tice dots were seen in this figure, which are labeled as A and B. These two dots were not similar in size. This phenomenon was also seen in the rest of the HRTEM sample observa‐ tions. The diameter of dot A is around 800 nm while that of dot B is around 700 nm. This is not agreeing with the LC-AFM image results. Through LC-AFM image observation, it is shown that the ZrO2 array dots are even and in the same diameter range of about 1 μm. This is because when the HRTEM sample was milling manually in the first stage, it was difficult to keep the milling plane vertical to the ZrO2 array plane. This is shown in Fig. 7(a) as the ideal case. It was always tilted a little bit which was perhaps due to the milling stage, sand‐ paper, or movement by hands. This is shown in Fig. 7(b) which is the practical case. For the conventional TEM specimen preparation techniques such as polished milling, the milling di‐ rection is controlled only by hands. Therefore, in the future HRTEM observation, an auto‐ matic monitor for sample milling system is necessary.

to the ZrO2 array plane. This is shown in Fig. 7(a) as the ideal case. It was always tilted a little bit which was perhaps due to the milling stage, sandpaper, or movement by hands. This is shown in Fig. 7(b) which is the practical case. For the conventional TEM specimen preparation techniques such as polished milling, the milling direction is controlled only by hands. This HRTEM image is accepted for interface analysis. In inset of Fig. 6(a), it shows the select‐ ed area diffraction pattern of platinum, ZrO2, and copper which are corresponding to poly‐ crystalline, amorphous, and polycrystalline behaviors, respectively. Fig. 6(a) shows the HRTEM image of copper layer and Fig. 6(c) shows the interface between ZrO2 array dot and platinum layer. In HRETM images, copper ions filaments were not found in the amorphous ZrO2 array dot. Therefore, it is possible that the Redox process could have happened in the interface between copper layer and ZrO2 array dot.

**Figure 6.** HRTEM interface image of Pt/ZrO2 lattice dot/Cu structure.

copper to form conducting copper-rich pathways. An opposite bias takes the existing copper

back to the copper electrode to its high resistance state [15-19].

conducting copper-rich pathways. An opposite bias takes the existing copper

Fig. 5. HRTEM image of Cu/ZrO2/ATO structure.

Fig. 6(a) shows the interface image of Pt/ZrO2 lattice dot/Cu sandwiched structure. Two lattice dots were seen in this figure, which are labeled as A and B. These two dots were not similar in size. This phenomenon was also seen in the rest of the HRTEM sample observations. The diameter of dot A is around 800 nm while that of dot B is around 700 nm. This is not agreeing with the LC-AFM image results. Through LC-AFM image observation, it is shown that the ZrO2 array dots are even and in the same diameter range of about 1 µm. This is because when the HRTEM sample was milling manually in the first stage, it was difficult to keep the milling plane vertical to the ZrO2 array plane. This is shown in Fig. 7(a) as the ideal case. It was always tilted a little bit which was perhaps due to the milling stage, sandpaper, or movement by hands. This is shown in Fig. 7(b) which is the practical case. For the conventional TEM specimen preparation techniques such as polished milling, the milling direction is controlled only by hands.

Fig. 6(a) shows the interface image of Pt/ZrO2 lattice dot/Cu sandwiched structure. Two lat‐ tice dots were seen in this figure, which are labeled as A and B. These two dots were not similar in size. This phenomenon was also seen in the rest of the HRTEM sample observa‐ tions. The diameter of dot A is around 800 nm while that of dot B is around 700 nm. This is not agreeing with the LC-AFM image results. Through LC-AFM image observation, it is shown that the ZrO2 array dots are even and in the same diameter range of about 1 μm. This is because when the HRTEM sample was milling manually in the first stage, it was difficult to keep the milling plane vertical to the ZrO2 array plane. This is shown in Fig. 7(a) as the ideal case. It was always tilted a little bit which was perhaps due to the milling stage, sand‐ paper, or movement by hands. This is shown in Fig. 7(b) which is the practical case. For the conventional TEM specimen preparation techniques such as polished milling, the milling di‐ rection is controlled only by hands. Therefore, in the future HRTEM observation, an auto‐

This HRTEM image is accepted for interface analysis. In inset of Fig. 6(a), it shows the select‐ ed area diffraction pattern of platinum, ZrO2, and copper which are corresponding to poly‐ crystalline, amorphous, and polycrystalline behaviors, respectively. Fig. 6(a) shows the HRTEM image of copper layer and Fig. 6(c) shows the interface between ZrO2 array dot and platinum layer. In HRETM images, copper ions filaments were not found in the amorphous ZrO2 array dot. Therefore, it is possible that the Redox process could have happened in the

back to the copper electrode to its high resistance state [15-19].

186 The Transmission Electron Microscope – Theory and Applications

**Figure 5.** HRTEM image of Cu/ZrO2/ATO structure.

matic monitor for sample milling system is necessary.

interface between copper layer and ZrO2 array dot.

**Figure 7.** HRTEM sample milling diagrams.

### **4. Conclusion**

Figure 7. HRTEM sample milling diagrams. **Acknowledgments** This work is supported by the National Natural Science Foundation of China (No. 51372198). The authors also thank the Foundation of Excellent Doctoral Dissertation of Xian University of Technology and the Scientific Research Projects of Shaanxi Education Department (No. 101-211111). **References** [1] B.J. Choi, D.S. Jeong, S.K. Kim, et al. Temperature influence and reset voltage study of bipolar resistive We fabricated ZrO2 thin films after annealing at 300°C through sol-gel deposition. Reprodu‐ cible I-V curves can be obtained with these samples at room temperature (300 K). The ZrO2 thin film annealed at 300°C in Cu/ZrO2/ATO device can also be operated in RS sweeps cy‐ cles. In summary, we fabricated TiO2 thin films by sol-gel deposition. Successful Cu/ ZrO2/ATO structure device in which ZrO2 thin film was calcined at 300°C can be obtained. The bipolar RS behavior was observed and the ratio of Roff:Ron can be reached to 104 . We also investigated the switching device of Pt/ZrO2 lattice dot/Cu sandwiched structure by HRTEM.

switching behaviour in ZrO2 thin films. *J Appl Phys*. 98 (2005), 037715-037725.

#### (2009), 628-632. [3] M.J. Rozenberg, I.H. Inoue, M.J. Sanchez. Nonvolatile memory with multilevel switching: A basic model. *Phys*  **Acknowledgements**

[4] X. Guo, C. Schindler, S. Menzel, R. Waser. Understanding the switching-off mechanism in Ag+ migration based resistively switching model systems. *Appl Phys Lett*. 91 (13) (2007), 133513-133516. [5] R. Waser. Resistive non-volatile memory devices. *Microelectron Eng*. 86 (2009), 1925-1928. [6] Y. Li, G. Zhao, J. Su, E. Shen, Y. Ren. Top electrode effects on 6 resistive switching behavior in CuO thin films. *Appl Phys A*. 104 (2011), 1069-1073. This work is supported by the National Natural Science Foundation of China (No. 51372198). The authors also thank the Foundation of Excellent Doctoral Dissertation of Xian University of Technology and the Scientific Research Projects of Shaanxi Education Department (No. 101-211111).

[7] C. Gopalan, M.N. Kozicki, S. Bhagat. *J Non-Crystal Solids*. 353 (2007), 1844-1848.

[2] H. Shima, Y. Tamai. Top electrode effects on resistive switching behavior in CuO thin films. *Microelectron J*. 40

[8] X. Zhi, G. Zhao, T. Zhu, Y. Li. The morphological, optical and electrical properties of SnO2:F thin films prepared

#### by spray pyrolysis. *Surf Interface Anal*. 40 (2008), 67-70. **Author details**

Ying Li1\*, Gaoyang Zhao2 , Zhibo Kou3 , Long Jin2 and Yajing Wang2

*Rev Lett.* 92 (17) (2004), 178302-178306.

\*Address all correspondence to: liying@xaut.edu.cn

1 Advanced Material Analysis Center, Xi'an University of Technology, Xi'an, Shaanxi, China

2 Material Science and Engineering School, Xi'an University of Technology, Xi'an, Shaanxi, China

3 Northwest Electric Power Design Institute of China Power Engineering Consulting Group, Xi'an, Shaanxi, China

### **References**

Figure 7. HRTEM sample milling diagrams.

188 The Transmission Electron Microscope – Theory and Applications

(2009), 628-632.

Shaanxi Education Department (No. 101-211111).

*Rev Lett.* 92 (17) (2004), 178302-178306.

*Appl Phys A*. 104 (2011), 1069-1073.

, Zhibo Kou3

\*Address all correspondence to: liying@xaut.edu.cn

This work is supported by the National Natural Science Foundation of China (No. 51372198). The authors also thank the Foundation of Excellent Doctoral Dissertation of Xian University of Technology and the Scientific Research Projects of

We fabricated ZrO2 thin films after annealing at 300°C through sol-gel deposition. Reprodu‐ cible I-V curves can be obtained with these samples at room temperature (300 K). The ZrO2 thin film annealed at 300°C in Cu/ZrO2/ATO device can also be operated in RS sweeps cy‐ cles. In summary, we fabricated TiO2 thin films by sol-gel deposition. Successful Cu/ ZrO2/ATO structure device in which ZrO2 thin film was calcined at 300°C can be obtained.

investigated the switching device of Pt/ZrO2 lattice dot/Cu sandwiched structure by

[1] B.J. Choi, D.S. Jeong, S.K. Kim, et al. Temperature influence and reset voltage study of bipolar resistive

[2] H. Shima, Y. Tamai. Top electrode effects on resistive switching behavior in CuO thin films. *Microelectron J*. 40

. We also

[3] M.J. Rozenberg, I.H. Inoue, M.J. Sanchez. Nonvolatile memory with multilevel switching: A basic model. *Phys* 

[4] X. Guo, C. Schindler, S. Menzel, R. Waser. Understanding the switching-off mechanism in Ag+ migration based

[6] Y. Li, G. Zhao, J. Su, E. Shen, Y. Ren. Top electrode effects on 6 resistive switching behavior in CuO thin films.

[8] X. Zhi, G. Zhao, T. Zhu, Y. Li. The morphological, optical and electrical properties of SnO2:F thin films prepared

and Yajing Wang2

switching behaviour in ZrO2 thin films. *J Appl Phys*. 98 (2005), 037715-037725.

The bipolar RS behavior was observed and the ratio of Roff:Ron can be reached to 104

resistively switching model systems. *Appl Phys Lett*. 91 (13) (2007), 133513-133516. [5] R. Waser. Resistive non-volatile memory devices. *Microelectron Eng*. 86 (2009), 1925-1928.

This work is supported by the National Natural Science Foundation of China (No. 51372198). The authors also thank the Foundation of Excellent Doctoral Dissertation of Xian University of Technology and the Scientific Research Projects of Shaanxi Education Department (No.

[7] C. Gopalan, M.N. Kozicki, S. Bhagat. *J Non-Crystal Solids*. 353 (2007), 1844-1848.

, Long Jin2

1 Advanced Material Analysis Center, Xi'an University of Technology, Xi'an, Shaanxi, China

2 Material Science and Engineering School, Xi'an University of Technology, Xi'an, Shaanxi,

3 Northwest Electric Power Design Institute of China Power Engineering Consulting Group,

by spray pyrolysis. *Surf Interface Anal*. 40 (2008), 67-70.

**Acknowledgments**

**Figure 7.** HRTEM sample milling diagrams.

**4. Conclusion**

HRTEM.

101-211111).

China

**Author details**

Ying Li1\*, Gaoyang Zhao2

Xi'an, Shaanxi, China

**References**

**Acknowledgements**


**Transmission Electron Microscopy for Biological Samples**

[16] C. Schindler, S.C.P. Thermadam, R. Waser, M.N. Kozicki. Bipolar and unipolar resis‐ tive switching in Cu-doped SiO2. *IEEE Trans Electr Dev*. 30 (1) (2007), 2762.

[17] A. Sawa, T. Fujii, M. Kawasaki. Interface resistance switching at a few nanometer

[18] K.M. Kim, B.J. Choi, Y.C. Shin, S. Choi, C.S. Hwang. Anode-interface localized fila‐ mentary mechanism in resistive switching of TiO2 thin films. *Appl Phys Lett*. 91

[19] C. Kügeler, R. Rosezin, E. Linn, R. Bruchhaus, R. Waser. Materials, technologies, and circuit concepts for nanocrossbar-based bipolar RRAM. *App*l Phys A. 102 (2011),

thick perovskite manganite active layers. *Appl Phys Lett*. 88 (2006),1-3.

(2007), 012907.

190 The Transmission Electron Microscope – Theory and Applications

791-809.

## **Transmission Electron Microscopy of Biological Samples**

Łukasz Mielańczyk, Natalia Matysiak, Olesya Klymenko and Romuald Wojnicz

Additional information is available at the end of the chapter

#### **Abstract**

During the last 70 years, transmission electron microscopy (TEM) has developed our knowledge about ultrastructure of the cells and tissues. Another aim is the determi‐ nation of molecular structure, interactions and processes including structure-function relationships at cellular level using a variety of TEM techniques with resolution in atomic to nanometre range. Even with the best transmission electron microscope, it is impossible to obtain real results without optimal sample preparation, respecting both the structure and the antigenicity preservation. Preparation techniques for highresolution study of both macromolecular complex and organelles within cellular complex are based on fast cryoimmobilisation process, where the sample is in the most native, hydrated state. Next, thin samples are directly visualised under cryotransmission electron microscopy (cryo-TEM), while thicker samples require a thinning step via cryo-electron microscopy of vitreous sections (CEMOVIS) or cryofocused ion beam (cryo-FIB) before visualisation. Alternatively, vitrified samples are freeze substituted and embedded in chosen resin for room temperature ultramicrot‐ omy. This preparation technique is suitable for morphological study, 3D analysis of cellular interior and immunoelectron microscopy. A different route for immunoloc‐ alisation study is cryosectioning according to the Tokuyasu technique that is a choice for rare or methacrylate-sensitive antigens. Most recently, new hybrid techniques have been developed for difficult-to-fix organisms and antigens or labile and anoxiasensitive tissues. Another preparation technique is, the oldest but still important, conventional chemical fixation dedicated in a wide range of research interest, involving morphological and immunolocalisation study. In this chapter, we present different sample preparation approaches for transmission electron microscopy of biological samples, including its methodological basis and applications.

© 2015 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

**Keywords:** conventional TEM, cryo-TEM, cryofixation, CEMOVIS, freeze substitu‐ tion, Tokuyasu technique, hybrid techniques, water

### **1. Introduction**

The first transmission electron microscope was constructed in the early 1930s by Ernst Ruska and Max Knoll [1]. Roughly a decade later, the first electron microscope picture of eukaryotic cells was taken by Keith Porter [2]. Since then TEM made it possible to study cells and tissue structure and function at nanoscale. This technique is placed between high-resolution methods like X-ray crystallography or nuclear magnetic resonance and lower-resolution light/fluores‐ cence microscopy techniques. Although fluorescence techniques allow for imaging dynamic process in living cells and modern fluorescence microscopes overcoming the diffraction limits that makes it possible to zoom in on cellular structure with resolution under 100 nm [3], TEM remains the main technique which makes it possible to study biological systems owing to its near-atomic-level resolution [4]. Moreover, TEM gives opportunities to visualise an interesting target with surrounding structure, when unlabelled surroundings still remain hidden at fluorescence sample [5]. Additionally, TEM comprises different branches: electron crystallog‐ raphy and single-particle analysis are dedicated to study proteins and macromolecular complexes, (cryo-)electron tomography and CEMOVIS for cellular organelles and molecular architectures and conventional TEM for gross morphology. Such a wide range of electron microscope techniques gives opportunity to find the relation between different macromole‐ cules, their supramolecular complexes and organelles assembled into an intricate network of cellular compartments. Knowledge of the cellular ultrastructure can contribute to an under‐ standing of how cells and tissues function in both normal physiological and pathological state.

Since the invention of the first TEM, the aim has been to image liquid samples at higher resolution but as easily as with light microscopy. At the beginning, it was impossible to accomplish due to low technological knowledge and lack of appropriate tools. Thus, scientists have introduced sample preparation techniques for observing soft and frail living matter in the inhospitable environment of an electron microscope. The TEM column is under ultrahigh vacuum, where electrons as a coherent beam are directed on the sample. From a biological point of view, living matter consists of up to 80% water; therefore, without appropriate sample preparation, the high vacuum literally sucks out every trace of liquid. Moreover, biological matter compose mainly of light elements (e.g. carbon, hydrogen, oxygen, nitrogen, sulphur, etc.) and for that reason, electrons which travel at a fraction of the speed of light, do not interact strongly with these atoms. Consequently, native biological materials are of extremely low contrast. On the other hand, the two aforementioned factors oriented sample preparation strategies. Due to the electron's scattering phenomena within the sample, only small objects can be observed directly. Larger samples need to be sectioned for analysis, but cells and tissues are too soft to be sectioned thinly enough without earlier sample preparation. Many labora‐ tories have been ingenious in designing and implementing different preparation techniques over the years, and as a result, scientists have found at least a partial remedy to these problems.

Therefore, a biological sample can be prepared either by removing or by freezing water. The oldest method is conventional sample preparation which uses chemical fixation, sample dehydration at room temperature and embedding with chosen resin. In the 1970s, Tokuyasu introduced an alternative to conventional sample preparation dedicated for immunocyto‐ chemistry. An alternative method to chemical fixation is cryofixation via vitrification process. Taking into consideration the size of the sample, electron microscopists have a wide range of freezing techniques at their disposal. Small or thin sample after plunge freezing can be directly observed at low temperature under cryo-transmission electron microscope (cryo-TEM). Thicker samples are first vitrified by using high-pressure freezing (HPF) or self-pressurised rapid freezing (SPRF) technique and then proceed to thinning process at liquid nitrogen temperature by CEMOVIS procedure or cryo-FIB milling. Finally, thin frozen-hydrated samples are directly observed under cryo-TEM. Another option is freeze substitution (FS) which bridges the gap between vitreous states and room temperature ultramicrotomy. Lastly, different combinations of mentioned techniques offer a new research possibility, especially for difficult-to-fix organisms or antigens. In a particular situation, chemical pre-fixation step is a prerequisite for successful sample vitrification, although it seems to be contradicted.

**Keywords:** conventional TEM, cryo-TEM, cryofixation, CEMOVIS, freeze substitu‐

The first transmission electron microscope was constructed in the early 1930s by Ernst Ruska and Max Knoll [1]. Roughly a decade later, the first electron microscope picture of eukaryotic cells was taken by Keith Porter [2]. Since then TEM made it possible to study cells and tissue structure and function at nanoscale. This technique is placed between high-resolution methods like X-ray crystallography or nuclear magnetic resonance and lower-resolution light/fluores‐ cence microscopy techniques. Although fluorescence techniques allow for imaging dynamic process in living cells and modern fluorescence microscopes overcoming the diffraction limits that makes it possible to zoom in on cellular structure with resolution under 100 nm [3], TEM remains the main technique which makes it possible to study biological systems owing to its near-atomic-level resolution [4]. Moreover, TEM gives opportunities to visualise an interesting target with surrounding structure, when unlabelled surroundings still remain hidden at fluorescence sample [5]. Additionally, TEM comprises different branches: electron crystallog‐ raphy and single-particle analysis are dedicated to study proteins and macromolecular complexes, (cryo-)electron tomography and CEMOVIS for cellular organelles and molecular architectures and conventional TEM for gross morphology. Such a wide range of electron microscope techniques gives opportunity to find the relation between different macromole‐ cules, their supramolecular complexes and organelles assembled into an intricate network of cellular compartments. Knowledge of the cellular ultrastructure can contribute to an under‐ standing of how cells and tissues function in both normal physiological and pathological state. Since the invention of the first TEM, the aim has been to image liquid samples at higher resolution but as easily as with light microscopy. At the beginning, it was impossible to accomplish due to low technological knowledge and lack of appropriate tools. Thus, scientists have introduced sample preparation techniques for observing soft and frail living matter in the inhospitable environment of an electron microscope. The TEM column is under ultrahigh vacuum, where electrons as a coherent beam are directed on the sample. From a biological point of view, living matter consists of up to 80% water; therefore, without appropriate sample preparation, the high vacuum literally sucks out every trace of liquid. Moreover, biological matter compose mainly of light elements (e.g. carbon, hydrogen, oxygen, nitrogen, sulphur, etc.) and for that reason, electrons which travel at a fraction of the speed of light, do not interact strongly with these atoms. Consequently, native biological materials are of extremely low contrast. On the other hand, the two aforementioned factors oriented sample preparation strategies. Due to the electron's scattering phenomena within the sample, only small objects can be observed directly. Larger samples need to be sectioned for analysis, but cells and tissues are too soft to be sectioned thinly enough without earlier sample preparation. Many labora‐ tories have been ingenious in designing and implementing different preparation techniques over the years, and as a result, scientists have found at least a partial remedy to these problems.

tion, Tokuyasu technique, hybrid techniques, water

194 The Transmission Electron Microscope – Theory and Applications

**1. Introduction**

Biological systems are very complex; thus, it is impossible to understand structure-function relationship outside the surrounding context. These days, dynamic developing of correlative light and electron microscopy (CLEM) approach can be observed. This approach relies on two steps. Firstly, the object of interest is located and imaged with fluorescence microscopy (FM), and then the sample is imaged in TEM. This technique is highly demanding according to cell biologists because high-resolution data can be fitted in the cellular context. However, new possibilities introduce new challenges in the preparative stage, and protocols for TEM and FM often are incompatible. Therefore, it is worth to mention that technological progress stimulates new sample preparation design, but often existing preparation schemes initiate new ideas.

In this chapter, we present different specialised preparation techniques dedicated to cells and tissues; but at the beginning, we would like to impress the importance of water in life on readers, because for a long time, its role in living organisms was neglected. Another point is that for a long time, water was treated as a foe by electron microscopists. Nevertheless, readers should bear in mind that selection of an appropriate technique strongly depends on the material and aims of the study. Thus, a general rule of thumb is that the higher the resolution is important, the closer to the native state sample preparation is desired. Moreover, the higher the resolution, the thinner sample should be, but at the same time, less information is achiev‐ able. During morphological study, more important is the sample size; hence, the preparation technique based on resin-embedded sample is an adequate choice. However, for immunolab‐ elling research, compromise between antigens and ultrastructure preservation is the major challenge. Although the main aim of this chapter is to present different preparation techniques of biological specimens for TEM, we would like to also point out that the preparation step is important for correlative approach.

We strongly encourage further reading of proposed positions where the reader can find practical insights of the presented subject, e.g. chosen volumes from *Methods in Molecular Biology* [6–8] and *Methods in Enzymology* [9–11], and with practical application in different model systems, positions from the Methods in Cell Biology series [12–14]. Many important hints in sample preparation for TEM are also connected with CLEM field [15,16] and immu‐ noelectron microscopy [17]. In our opinion, a complete library should also include the *Handbook of Cryo-Preparation Methods for Electron Microscopy* because this position is strongly oriented to the practical side of sample preparation art [18]. It is also important to know what was done so far and thus where we should go. Among many old books, but with still-current knowledge, *Cryotechniques in Biological Electron Microscopy* [19] captured our attention. The last but not the least position is the *Principles and Techniques of Electron Microscopy: Biological Applications* [20].

### **2. Water and its vital role in life**

Organisms consist in major part of liquid water which is the medium in which life takes place. Hence, life on our planet and its probability elsewhere in the universe cannot have evolved or continue without water. In view of the abundant presence of water in living organisms, this substance cannot be perceived as an inert diluent. Water performs many functions: it trans‐ ports, reacts, lubricates and structures and is used in signalling. Water is also a metabolite and a temperature buffer. The physical properties of water, which result from its structure, play a key role in the orchestration of the cell machinery. Biological molecules and water should be thought as equal partners where one is required and structured by the other.

From a chemical point of view, a water molecule contains one oxygen atom covalently bounded with two hydrogen atoms. Due to positively charged hydrogen atoms and a negatively charged oxygen atom, where negative charge comes from two lone electron pairs, water is a dipole. Water as a dipole has the most important property: water molecules are able to form multiple hydrogen bonds between each other. A hydrogen bond occurs when a partially positively charged hydrogen atom lies between partially negatively charged oxygen of H2O molecules. A hydrogen bond is naturally formed from a complex combination of different interactions: an electrostatic, a polarisation and a covalent attraction, and a dispersive attrac‐ tive interactions, an electron repulsion and a nuclear quantum effects. In theory, one water molecule can interact with four other water molecules, thereby forming a tetrahedron config‐ uration. In practice, hydrogen bonds are very dynamic and heterogeneous structures, both on energetic and structural levels, and a single water molecule can form two or four hydrogen bonds. As a result, in liquid water, hydrogen bonds behave in cooperative and anticooperative manner [21]. At the higher level of organisation, water molecules in liquid state tend to create tetrahedral pentameric clusters, which are linked to other water molecules and clusters to form a complex network or liquid phases [22]. Such a network of hydrogen bonds is dynamic and ordered in a nanometre range structure.

From a biological point of view, a cell's interior consists of membranes and cytoskeleton together with proteins, carbohydrates and nucleic acids which are strongly and inseparable dependent on water molecules. The dipolar nature of water enables to arrange molecules of water into an ordered, very constrained manner on the surface of biological molecules. Depending on the chemical nature of surface domains, hydrophobic or hydrophilic, water order is different. Water molecules are strongly attracted by ionised and hydrophilic domains than by apolar domains, where H2O molecules arrange themselves into clathrate-like struc‐ tures [23]. They form a hydration shell, called also interfacial water, built from several water layers. Hydration shells are critical for solubility of molecules and prevent them from aggre‐ gating. When two particles meet, they do not stick together, but separate [24]. Moreover, protein folding is mediated and guided by aqueous solvation, and protein structure is stabilised by water clusters and their hydrogen bonding capabilities. Water also gives proteins flexibility during conformational changes, and its molecules mediated protein-ligand interac‐ tion. Another interesting example of water role in the cellular world is nucleic acid-water interactions. Firstly, water molecules stabilise structure of double helix. Secondly, water hydrates both the major and the minor grooves of DNA. The hydration shells have a charac‐ teristic pattern reliant on bases and their sequences and thus create a 'hydration fingerprint' for a given DNA sequence. The specific arrangement of interfacial water governs protein binding to the DNA. The enumerated examples are further discussed in detail in [25]. Water inside the cell, which is not bound in hydration shell, is unaffected by the biomolecules. Additionally, cellular unbounded water behaves differently from water outside, e.g. intracel‐ lular water has higher viscosity. The cytoplasm has a sol-gel nature. The local parts of the cytoplasm may manifest itself as a more highly viscous and stiff environment, likened to a gelstate, or as a low-viscosity sol-state solution [21]. In the former case, water molecules form more strongly hydrogen-bounded water clusters. This reduces local fluctuations in the nearby macromolecules and slows down metabolite and ion migration [23]. An additional function of strong hydrogen-bonded network existence is transmission of information about solutes and surfaces at distances of several nanometres. In this way, biomolecules 'sense' each other, thereby changing their solubility and activity. In the latter case, proteins can release a signifi‐ cant amount of water, changing the fluidity and activity of the intracellular 'sol'-state envi‐ ronment. The state of the water is thus essential for the biological activity of the cell, and the state of metabolites controls water structure. Thus, water is defined as an engine of life [26].

From the physical point of view, water is usually perceived as an ordinary substance because people interact with it all the time in their everyday lives. Very often, people think that other liquids have similar properties to this liquid; however, nothing could be further from the truth. Water is the only liquid that exists in all three states on Earth: liquid, solid and gas. Although some researchers believe that intracellular water is the fourth phase of water [27]. Some properties of water, such as large heat capacity and high thermal conductivity, allow the control of body temperature. The high latent heat of evaporation is a protection from dehy‐ dration and considerable evaporative cooling [21]. To sum up, water is essential for life existence at different levels: from molecular to cellular and organisms to whole-planet level. All properties of this substance are not yet known; therefore, new discoveries in this field will have an interesting impact in understanding how life works.

#### **3. Conventional TEM**

noelectron microscopy [17]. In our opinion, a complete library should also include the *Handbook of Cryo-Preparation Methods for Electron Microscopy* because this position is strongly oriented to the practical side of sample preparation art [18]. It is also important to know what was done so far and thus where we should go. Among many old books, but with still-current knowledge, *Cryotechniques in Biological Electron Microscopy* [19] captured our attention. The last but not the least position is the *Principles and Techniques of Electron Microscopy: Biological Applications* [20].

Organisms consist in major part of liquid water which is the medium in which life takes place. Hence, life on our planet and its probability elsewhere in the universe cannot have evolved or continue without water. In view of the abundant presence of water in living organisms, this substance cannot be perceived as an inert diluent. Water performs many functions: it trans‐ ports, reacts, lubricates and structures and is used in signalling. Water is also a metabolite and a temperature buffer. The physical properties of water, which result from its structure, play a key role in the orchestration of the cell machinery. Biological molecules and water should be

From a chemical point of view, a water molecule contains one oxygen atom covalently bounded with two hydrogen atoms. Due to positively charged hydrogen atoms and a negatively charged oxygen atom, where negative charge comes from two lone electron pairs, water is a dipole. Water as a dipole has the most important property: water molecules are able to form multiple hydrogen bonds between each other. A hydrogen bond occurs when a partially positively charged hydrogen atom lies between partially negatively charged oxygen of H2O molecules. A hydrogen bond is naturally formed from a complex combination of different interactions: an electrostatic, a polarisation and a covalent attraction, and a dispersive attrac‐ tive interactions, an electron repulsion and a nuclear quantum effects. In theory, one water molecule can interact with four other water molecules, thereby forming a tetrahedron config‐ uration. In practice, hydrogen bonds are very dynamic and heterogeneous structures, both on energetic and structural levels, and a single water molecule can form two or four hydrogen bonds. As a result, in liquid water, hydrogen bonds behave in cooperative and anticooperative manner [21]. At the higher level of organisation, water molecules in liquid state tend to create tetrahedral pentameric clusters, which are linked to other water molecules and clusters to form a complex network or liquid phases [22]. Such a network of hydrogen bonds is dynamic and

From a biological point of view, a cell's interior consists of membranes and cytoskeleton together with proteins, carbohydrates and nucleic acids which are strongly and inseparable dependent on water molecules. The dipolar nature of water enables to arrange molecules of water into an ordered, very constrained manner on the surface of biological molecules. Depending on the chemical nature of surface domains, hydrophobic or hydrophilic, water order is different. Water molecules are strongly attracted by ionised and hydrophilic domains than by apolar domains, where H2O molecules arrange themselves into clathrate-like struc‐

thought as equal partners where one is required and structured by the other.

**2. Water and its vital role in life**

196 The Transmission Electron Microscope – Theory and Applications

ordered in a nanometre range structure.

As the vital role of water was outlined in the last paragraph, it must be borne in mind that the knowledge about water and its role in living organisms has evolved over decades. Recently, however, the role of water in cell and molecular biology has become clearer and much more important than in the past. At the beginning of the sample preparation, biologists carefully eliminate every trace of water. Conventional specimen preparation is the most commonly used method for TEM. The major advances in conventional specimen preparation technique are summarised at the end of the section.

The first step in this procedure is chemical fixation to preserve the biological sample with minimal alteration of volume and morphology from the native state. Chemical fixation can be carried out in various fixatives which are used separately or in combination, i.e. glutaraldehyde (GA), paraformaldehyde, osmium tetroxide, uranyl acetate (UA) and tannic acid solution are usually used. Glutaraldehyde, as a dialdehyde, preserves ultrastructure well but penetrates slower than the monoaldehyde, i.e. paraformaldehyde. Glutaraldehyde is used alone for small pieces of material, but a mixture of the two aldehydes may be used for fixation of larger items. The most popular fixation strategy in conventional preparation is double fixation with GA and osmium tetroxide. The aldehydes preserve mainly proteins, but reaction with lipids is limited. Therefore, to stabilise the lipid part of the sample, the post-fixation with osmium tetroxide is required [28]. Furthermore, osmication enhances contrast which is important during the analysis in the TEM, but it is important to remember that prolonged process can destroy proteins; as a consequence, the biological material becomes brittle. It should also be noted that the common practice in conventional electron microscopy is en bloc fixation/staining with UA and tannic acid. The first fixative decreases protein and phospholipid extraction; the second one reveals ultrastructure of microtubules.

The chemical fixation depending on its sample type and size can be performed in four various ways: by in situ fixation (applied to cell cultures), by immersion (small pieces of tissues are carefully excised and immersed in fixative as soon as possible), by vascular perfusion (the fixative is pumped through the vascular system of deeply sedated animals) or by vapours (the small delicate specimens, such as membranes, are suspended over a solution of osmium tetroxide). Usually, in standard conditions, fixation is carried out by immersion method. After fixation with buffered fixatives, the sample is dehydrated in increasing concentration of a solvent (a combination of either alcohol or acetone with propylene oxide) to enable infiltration with a liquid resin. An epoxy resin is most commonly used in conventional TEM. Finally, supersaturated and surrounded by resin, the sample is polymerised by applying gentle heat. Cured resin block containing biological material is thinly sectioned (40–150 nm), and subse‐ quently post-stained with heavy metal salts, such as uranyl, and lead in order to introduce contrast inside the sample.

Previous reports revealed that chemical fixation, dehydration, heavy metal staining and plastic embedding can introduce various artefacts. Fixation with GA prior to dehydration results in cross-linking, causing aggregation of proteins, collapse of highly hydrated glycans and loss of lipids. Heavy metals can cause additional artefacts in the form of precipitation [24,29,30]. To sum up, artefacts introduced during conventional sample preparation limit the resolution of biological samples to about 2 nm [20]. Further, modern transmission electron microscopes with a field emission gun can obtain sub-angstrom resolution; thus, the resolution is mainly limited by the properties of the sample [31]. Therefore, the improvement of biological sample prepa‐ ration technique has become the challenge for many scientists, and nowadays, some interesting alternatives are available.

On the other side, it should be stressed that TEM with conventional preparation is still an essential tool in many fields of tissue and cell biology, as well as in medicine. Where the approach is quality control method or gold standard to complement, support or confirm the results of specific histopathological diagnosis (neoplastic, renal, neuromuscular, infectious, hereditary and metabolic diseases) [32,33]. Therefore, the advantages of the conventional sample preparation should be emphasised, starting from simplicity of this fixation technique. There is no requirement for any specialised equipment; indeed, simple vessels with fixatives are sufficient. At the same time, numerous 1-mm3 -in-volume blocks of tissue can be fixed; thus, large areas of sample are accessible to analysis. When the tissue autolyses quickly, organs are too large or dissecting is difficult, chemical fixation is carried out by perfusion. Appropriate optimisation of fixation parameters, i.e. fixative concentration, buffer pH, temperature of fixative and time fixation, results in optimal ultrastructure preservation. Therefore, the unique advantage of conventional fixation is its ability to fix human tissue biopsies and study different animal organs without biopsy need. Therefore, the improvement of biological sample preparation technique has become the challenge for many scientists, and nowadays, some interesting alternatives are available. On the other side, it should be stressed that TEM with conventional preparation is still an essential tool in many fields of tissue and cell biology, as well as in medicine. Where the approach is quality control method or gold standard to complement, support or confirm the results of specific histopathological diagnosis (neoplastic, renal, neuromuscular, infectious, hereditary and metabolic diseases) [32,33]. Therefore, the advantages of the conventional sample preparation should be emphasised, starting from simplicity of this fixation technique. There is no requirement for any specialised equipment; indeed, simple vessels with fixatives are sufficient. At the same time, numerous 1-mm3 -in-volume blocks of tissue can be fixed; thus, large areas of sample are accessible to analysis. When the tissue autolyses quickly, organs are too large or dissecting is difficult, chemical fixation is carried out by perfusion. Appropriate optimisation of fixation parameters, i.e. fixative concentration, buffer pH, temperature of fixative and time fixation, results in optimal ultrastructure preservation. Therefore, the unique advantage of conventional fixation is its ability to fix human

important than in the past. At the beginning of the sample preparation, biologists carefully eliminate every trace of water. Conventional specimen preparation is the most commonly used method for TEM. The major advances in conventional specimen preparation technique are

The first step in this procedure is chemical fixation to preserve the biological sample with minimal alteration of volume and morphology from the native state. Chemical fixation can be carried out in various fixatives which are used separately or in combination, i.e. glutaraldehyde (GA), paraformaldehyde, osmium tetroxide, uranyl acetate (UA) and tannic acid solution are usually used. Glutaraldehyde, as a dialdehyde, preserves ultrastructure well but penetrates slower than the monoaldehyde, i.e. paraformaldehyde. Glutaraldehyde is used alone for small pieces of material, but a mixture of the two aldehydes may be used for fixation of larger items. The most popular fixation strategy in conventional preparation is double fixation with GA and osmium tetroxide. The aldehydes preserve mainly proteins, but reaction with lipids is limited. Therefore, to stabilise the lipid part of the sample, the post-fixation with osmium tetroxide is required [28]. Furthermore, osmication enhances contrast which is important during the analysis in the TEM, but it is important to remember that prolonged process can destroy proteins; as a consequence, the biological material becomes brittle. It should also be noted that the common practice in conventional electron microscopy is en bloc fixation/staining with UA and tannic acid. The first fixative decreases protein and phospholipid extraction; the second

The chemical fixation depending on its sample type and size can be performed in four various ways: by in situ fixation (applied to cell cultures), by immersion (small pieces of tissues are carefully excised and immersed in fixative as soon as possible), by vascular perfusion (the fixative is pumped through the vascular system of deeply sedated animals) or by vapours (the small delicate specimens, such as membranes, are suspended over a solution of osmium tetroxide). Usually, in standard conditions, fixation is carried out by immersion method. After fixation with buffered fixatives, the sample is dehydrated in increasing concentration of a solvent (a combination of either alcohol or acetone with propylene oxide) to enable infiltration with a liquid resin. An epoxy resin is most commonly used in conventional TEM. Finally, supersaturated and surrounded by resin, the sample is polymerised by applying gentle heat. Cured resin block containing biological material is thinly sectioned (40–150 nm), and subse‐ quently post-stained with heavy metal salts, such as uranyl, and lead in order to introduce

Previous reports revealed that chemical fixation, dehydration, heavy metal staining and plastic embedding can introduce various artefacts. Fixation with GA prior to dehydration results in cross-linking, causing aggregation of proteins, collapse of highly hydrated glycans and loss of lipids. Heavy metals can cause additional artefacts in the form of precipitation [24,29,30]. To sum up, artefacts introduced during conventional sample preparation limit the resolution of biological samples to about 2 nm [20]. Further, modern transmission electron microscopes with a field emission gun can obtain sub-angstrom resolution; thus, the resolution is mainly limited by the properties of the sample [31]. Therefore, the improvement of biological sample prepa‐ ration technique has become the challenge for many scientists, and nowadays, some interesting

summarised at the end of the section.

198 The Transmission Electron Microscope – Theory and Applications

one reveals ultrastructure of microtubules.

contrast inside the sample.

alternatives are available.

tissue biopsies and study different animal organs without biopsy need.

**Figure 1a** Example of membranous glomerulopathy. Electron photomicrograph shows large amounts of electron-dense deposits in the glomerular basement membrane (black arrows); the sub-epithelial deposits are covered by a bridge of newly formed glomerular base membrane (white arrowheads). The foot processes of the epithelial cells are obliterated. In the cytoplasm of the epithelial and endothelial cells are numerous organelles and vacuoles. **Figure 1b** Electron micrograph chemically fixed and an Epon-embedded rat's small intestine. Ultrastructure of the brush border (BB) and **Figure 1. (a)** Example of membranous glomerulopathy. Electron photomicrograph shows large amounts of electrondense deposits in the glomerular basement membrane (black arrows); the sub-epithelial deposits are covered by a bridge of newly formed glomerular base membrane (white arrowheads). The foot processes of the epithelial cells are obliterated. In the cytoplasm of the epithelial and endothelial cells are numerous organelles and vacuoles. **(b)** Electron micrograph chemically fixed and an Epon-embedded rat's small intestine. Ultrastructure of the brush border (BB) and terminal web (TW) region of an enterocyte. Notice the actin filaments (white arrows) that descend from each microvil‐ lus deep into the underlying terminal web; below, organelles are visible. Lower part: transverse section through brush border showing numerous microvilli containing actin filaments (white arrow) and fusion of microvilli (black arrows).

from each microvillus deep into the underlying terminal web; below, organelles are visible. Lower In particular cases, pre-fixation with aldehydes is a prerequisite for further sample processing (see Sect. 8). Another advantage of chemical fixation is the possibility to perform sample preparation in different locations, e.g. in an operating room, in a laboratory other than EM lab and in the natural environment.

terminal web (TW) region of an enterocyte. Notice the actin filaments (white arrows) that descend

Adding to this knowledge about artefacts that are introduced during sample preparation, these advantages constitute strong position of conventional TEM in modern science. To prove this statement, some examples are presented. The conventional sample preparation in ultrastruc‐ tural pathology is often irreplaceable (Figure 1a), as evidenced by different books [32,34]. A human biopsy from a patient with mitochondrial cardiomyopathy may serve as another example, where pathogenic giant mitochondria probably are formed to compensate effects caused by mutation of mitochondrial DNA [35]. Also, some interesting discoveries at scientific ground are contributed by conventional TEM. Among examples, electron microscopic analysis of a spherical structure of mitochondria is quite interesting [36]. Conventional sample preparation combined with ET revealed that mitochondria under oxidative stress were able to undergo a structural transformation in spheroid form. This novel mitochondrial dynamic process is probably involved in some pathological conditions; however, further study is needed.


Telocytes are a new type of interstitial cells characterised by the presence of telopodes, visible only by TEM [37]. These cells can be found in most organs, including the heart. The recent study showed the new type of extracellular vesicles released by heart telocytes that probably are an essential component in the paracrine secretion and may be involved in the heart physiology and regeneration process [38]. The combination of classical TEM and cell culture is an unbiased approach to identify unrecognised pathogen, provide the first clue about investigated pathogen and guide further laboratory study and epidemiology [39]. Another example is ultrastructure of the glomerular filtration barrier in the kidney [40]. ET in combi‐ nation with scanning electron microscopy (SEM) confirmed suggestions that the glomerular filtration barrier comprises five layers instead of three. This discovery should allow for deeper understanding of kidney physiology and diseases.

Although conventional TEM is also appropriate for immunogold labelling, we leave this approach out. For more information about immunogold labelling on resin section, refer to Chapter **X**.

### **4. The Tokuyasu technique**

Adding to this knowledge about artefacts that are introduced during sample preparation, these advantages constitute strong position of conventional TEM in modern science. To prove this statement, some examples are presented. The conventional sample preparation in ultrastruc‐ tural pathology is often irreplaceable (Figure 1a), as evidenced by different books [32,34]. A human biopsy from a patient with mitochondrial cardiomyopathy may serve as another example, where pathogenic giant mitochondria probably are formed to compensate effects caused by mutation of mitochondrial DNA [35]. Also, some interesting discoveries at scientific ground are contributed by conventional TEM. Among examples, electron microscopic analysis of a spherical structure of mitochondria is quite interesting [36]. Conventional sample preparation combined with ET revealed that mitochondria under oxidative stress were able to undergo a structural transformation in spheroid form. This novel mitochondrial dynamic process is probably involved in some pathological conditions; however, further study is

**Key events in the history of conventional TEM in the life science**

1932–1934 The first images of unfixed biological material were obtained (the wings and legs of a housefly)

<sup>1949</sup> Newman et al. introduced methacrylates which became the first embedding media of

1952 Palade recommended buffered osmium fixative for the preservation of cell ultrastructure

1956–1982 Introduction of new resins for electron microscopy: Araldite (1958), epoxy resin (1961)

1959 Singer employed ferritin coupled with immunoglobulins to recognise the location of the antigen of interest

1966 Graham and Karnovsky developed a cytochemical method for horseradish peroxidise localisation

1971 Faulk and Taylor introduced colloidal gold labelling technique for ultrathin sections

David D. Sabatini introduced the aldehyde fixation; previously used fixatives were replaced by double fixation with glutaraldehyde and osmium tetroxide. This fixation strategy revolutionised the field of biological electron microscopy, and it is still the method of choice for many laboratories nowadays

Fernandez-Moran used for ultrathin sectioning a diamond knife Potassium permanganate fixation was introduced by Luft

<sup>1943</sup> Sjöstrand built the first ultramicrotome and developed a method of producing ultrathin tissue samples which were used to study the skeletal muscle. Rapid development of ultramicrotomy since 1948 1945 Porter, Claude and Fullam imaged whole eukaryotic cells that were fixed in osmium vapour and then dried

1932 Ruska and Knoll built the first transmission electron microscope

1939 Ruska obtained the first bacterium and virus TEM pictures

200 The Transmission Electron Microscope – Theory and Applications

1950 Latta and Hartmann used glass knife in ultramicrotomy

1958 Kellenberger used UA to stabilise viral and bacterial DNA

1963 Reynolds used lead citrate as electron-dens stain in TEM

quality for TEM samples

and Lowicryl media (1982)

needed.

1956

1963

Cryosectioning according to Tokuyasu is one of the most reliable and sensitive immunolocal‐ isation techniques for different types of sample. It was introduced in the 1970s by Tokuyasu, using a cryo-ultramicrotome developed by A. Christensen [41], although the first attempts at cryosectioning were pioneered by Fernandez-Moran in 1952 [42] and further developed by Bernhard and Leduc [43]. Similarly to the conventional specimen preparation techniques, the biological material is first chemically fixed with aldehydes at low concentration. Instead of dehydration with organic solvent and resin embedding, the fixed material is infiltrated with sucrose, subsequently frozen in liquid nitrogen and sectioned with a dry knife at low temper‐ ature. What is interesting is in the past, it was suggested that wet sectioning based on water mixtures with sucrose, dimethyl sulfoxide or glycerol may resemble conventional resin ultramicrotomy. This idea dies a natural death because of limited temperature range for optimisation of the cryosectioning and chemical properties of proposed mixtures [44]. Obtained cryosections are retrieved from the knife surface, thawed and placed either onto microscope slides for light microscopy applications or electron microscopy formvar-carboncoated grids. Such prepared cryosections are ready to perform immunolocalisation step on both fluorescence microscopy [45] and EM level [46]. For the latter, immunolabelled cryosec‐ tions are stained with UA, embedded in methylcellulose to prevent drying and cell organelle irregular shrinkage artefacts and examined in the TEM [41,47]. The final step for fluorescence microscopy is mounting the coverslip with immunolabelled material on a slide with a drop of mounting medium before examination.

At a recent time, a new and improved contrasting procedure called 2+Staining was introduced for Tokuyasu cryosections for correlative approach [48]. 2+Staining consists of contrasting immunolabelled sections with 1% osmium tetroxide, 2% UA and lead citrate and followed by embedding in 1.8% methylcellulose. In comparison with the Griffith method, where membrane contrast is excellent but fluorescence signal on the section is strongly reduced or lost altogether [49], the introduced procedure yielded a positive contrast of the cellular organelles and membranes, as in the contrast obtained from stained resin-embedded sections. The 2+Staining procedure shows little to no effect on the signal of the fluorescence labelling on the sections.

In comparison with conventional technique, the sucrose infiltration step dehydrates chemically fixed specimen in lesser degree. Thus, the advantage of thawed cryosection labelling over resin sections stems from the preservation of a natural hydrophilic environ‐ ment. Moreover, resin sections are created with a polymeric cross-linked matrix; thus, antigen accessibility is lower and mainly restricted to the resin surface. In consequence, thawed cryosections enable better access to the antigens, particularly rare ones, for the probes and high detection sensitivity [50]. Over the years, the technique has been further improved by Tokuyasu and his colleagues [51–54]. These improvements become basis for the future research on various biological organisms and hybrid technique developments [48,55–59] including correlative approach [4,48].

### **5. Vitrification as an alternative to conventional fixation**

An alternative way to deal with water is keep it inside the biological specimen by cryofixation. Cryofixation is based on vitrification that is not fixation per se, because vitrification is a physical process of the solidification of a liquid into noncrystalline or amorphous solid known as glass, usually induced by rapid cooling [60]. Water molecules do not have time to start to crystallise because heat is extracted from the system with high efficiency. Accordingly, term fixation cannot be understood in a chemical sense, where covalent bond is created. In the case of water, its vitreous state can be seen as liquid water with extremely high viscosity. Hence, cryofixation of biological matter allows instantaneous immobilisation of all cell constituents in their current positions which translates to obtain a true snapshot of the cell at the moment of ultrarapid freezing [61]. To précis, diffusible ions and molecular components are not shifted or extracted from the cell [62–64]; ultrastructural morphology is preserved close to the nature state, and rapid physiological processes can be characterised in a precise point in time.

When cooling rate is not sufficient, the formation of crystalline water destroys cellular ultrastructure. Another anomalous property of water is its increasing volume during crystal‐ lisation. Water freezing is not the reverse process of ice melting [65]. The freezing process is complex and influenced by nucleation centre, crystal growth and cooling rate. Slow cooling rate at ambient pressure results in hexagonal ice, whereas cubic ice forms under faster cooling rate, especially in the presence of nucleation centre. However, liquid water below its melting point is supercooled water. Liquid water is easily supercooled down to about -25 ºC and in 9 μm-diameter tiny droplet, down to about -46 ºC [66] and -92 ºC at 204,8 MPa [60]. Below these temperatures, crystal formation rapidly takes place as supercooled water has to go through the nucleation process by itself, generating nucleation centre distributed uniformly over the phase. In the end, ice crystals are formed. In the context of biological matter, cellular interior is a complex solution of different soluble materials, such as ions, proteins and sugars that interact with the water molecules, thus changing water freezing behaviour. The presence of salts and hydrophilic solutes increase the extent of supercooling by lowering the freezing point of the solution. Furthermore, increasing solution concentration increases the temperature at which water becomes vitreous. In general, there is lack of heterogeneous nucleation centres inside the cell, and cellular solution concentration is usually higher compared to the sur‐ rounding medium. For these reasons, freezing starts somewhere in the external medium or on the surface where cells or tissues are placed. Growing ice crystals outside the cell are avid for any water molecules, including intracellular water, resulting in osmotic pressure changes and eventually bursting of the membrane [67]. Finally, biological matter after freezing consists of ice crystals and dehydrated solution between ice ramifications. Therefore, the water in the cell and around it should always be vitrified to avoid deleterious effects of ice formation. The mentioned issue underlies the success of the Tokuyasu method, where cryoprotected chemi‐ cally fixed sample is vitrified by dropping it into liquid nitrogen. Nevertheless, the time when vitrification becomes more understandable and easier to perform had yet to come.

The preparation of biological samples by vitrification process is relatively a new technique. Until 1980, it was thought that water cannot be vitrified because of its thermodynamic properties. However, the big breakthrough came in 1980 when Mayer and Bruggeler had used X-ray diffraction to prove that small water droplets can be vitrified [68]. One year later, Dubochet and McDowall published an article where they demonstrated that small droplets of water or aqueous solution can be vitrified on electron microscopic grids [69]. These discoveries marked a turning point in biology, especially structural biology, and changed the point of view on water in electron microscopy from being the foe to being the best friend of electron microscopists [70]. The vitrification process depends on biological and physical factors such as the thermal diffusivity, the thickness and solute concentration of the sample as well as the cooling rate and the pressure applied. Nonetheless, the cooling rate seems to be the most important factor and decides which technique should be used [60]. Different preparation techniques are available for different sample sizes and the microscopy approach.

#### **5.1. Plunge freezing**

contrast is excellent but fluorescence signal on the section is strongly reduced or lost altogether [49], the introduced procedure yielded a positive contrast of the cellular organelles and membranes, as in the contrast obtained from stained resin-embedded sections. The 2+Staining procedure shows little to no effect on the signal of the fluorescence labelling on the sections.

In comparison with conventional technique, the sucrose infiltration step dehydrates chemically fixed specimen in lesser degree. Thus, the advantage of thawed cryosection labelling over resin sections stems from the preservation of a natural hydrophilic environ‐ ment. Moreover, resin sections are created with a polymeric cross-linked matrix; thus, antigen accessibility is lower and mainly restricted to the resin surface. In consequence, thawed cryosections enable better access to the antigens, particularly rare ones, for the probes and high detection sensitivity [50]. Over the years, the technique has been further improved by Tokuyasu and his colleagues [51–54]. These improvements become basis for the future research on various biological organisms and hybrid technique developments

An alternative way to deal with water is keep it inside the biological specimen by cryofixation. Cryofixation is based on vitrification that is not fixation per se, because vitrification is a physical process of the solidification of a liquid into noncrystalline or amorphous solid known as glass, usually induced by rapid cooling [60]. Water molecules do not have time to start to crystallise because heat is extracted from the system with high efficiency. Accordingly, term fixation cannot be understood in a chemical sense, where covalent bond is created. In the case of water, its vitreous state can be seen as liquid water with extremely high viscosity. Hence, cryofixation of biological matter allows instantaneous immobilisation of all cell constituents in their current positions which translates to obtain a true snapshot of the cell at the moment of ultrarapid freezing [61]. To précis, diffusible ions and molecular components are not shifted or extracted from the cell [62–64]; ultrastructural morphology is preserved close to the nature state, and

When cooling rate is not sufficient, the formation of crystalline water destroys cellular ultrastructure. Another anomalous property of water is its increasing volume during crystal‐ lisation. Water freezing is not the reverse process of ice melting [65]. The freezing process is complex and influenced by nucleation centre, crystal growth and cooling rate. Slow cooling rate at ambient pressure results in hexagonal ice, whereas cubic ice forms under faster cooling rate, especially in the presence of nucleation centre. However, liquid water below its melting point is supercooled water. Liquid water is easily supercooled down to about -25 ºC and in 9 μm-diameter tiny droplet, down to about -46 ºC [66] and -92 ºC at 204,8 MPa [60]. Below these temperatures, crystal formation rapidly takes place as supercooled water has to go through the nucleation process by itself, generating nucleation centre distributed uniformly over the phase. In the end, ice crystals are formed. In the context of biological matter, cellular interior is a complex solution of different soluble materials, such as ions, proteins and sugars that

[48,55–59] including correlative approach [4,48].

202 The Transmission Electron Microscope – Theory and Applications

**5. Vitrification as an alternative to conventional fixation**

rapid physiological processes can be characterised in a precise point in time.

The simplest way to obtain a vitrified sample is through the so-called bare grid method [71] or plunge freezing (PF). Typically, vitrification is done by an immersion of small biological objects within a thin water film into properly chosen cryogen. After vitrification, the sample is directly imaged in a frozen state under a cryo-electron microscope. Under atmospheric pressure, cooling rate as high as 108 K/s is possible, at least in theory [72]. The successful vitrification obtained by PF is strongly dependent on properties of the grid, the temperature and nature of cryogen and the environment of the sample. The cryogen of choice is ethane, cooled to their freezing points by thermal contact with liquid nitrogen [71]. However, the usage of the mixture composed of propane and ethane has two advantages over pure ethane. Firstly, pure alkane is solid at 77 K, whereas the above mixture remains as a liquid at this temperature. For that reason, ethane-propane mixture does not require repeated cooling and warming cycles in order to ensure proper vitrification conditions. Secondly, the mixture produces less damage to carbon layer on the grids and to specimens mounted on C-flat holey carbon grids [73]. In addition, the surrounding environment has an influence on the success of the final vitrification. In the case of thin films, usually 100 nm thick, prepared before plunging, sample is subjected to evaporation and heat transfer processes. In consequence, sample is exposed to risk of concentration, pH, ionic strength and temperature changes. Therefore, a new method was devised by Bellare et al. [74] and further improved by Frederik [75], who has constructed a new robot for sample vitrification by this technique. Plunge freezing can be applied to viruses [76], bacteria [77,78], isolated protein complexes [79,80] and whole cells [81,82], however, only 10 μm thick. In the last case, the flat thin periphery of cells [81] or whole thin cells, e.g. *Ostreococcus* [83] and *Plasmodium* sporozoites [84], is directly observed.

Although high cooling rate values are obtained during vitrification by PF, this technique is suitable for samples as thick as few micrometres. The reason is poor heat conduction properties of water because even an infinitely high cooling rate applied at the sample surface rapidly decays within the sample. Thicker-sample vitrification is only possible when cooling rate is reduced under particular conditions [60,85].

#### **5.2. High-pressure freezing**

The high-pressure freezing (HPF) technique was developed in the 1960s by Moor and Riehle [86], but as a routine laboratory technique, it became available from the mid-1980s. This freezing technique is based on Le Chatelier's principle and on the fact that physical properties of water can be manipulated in some range. Water during crystallisation expands its volume and forms crystal. According to Le Chatelier's principle, the increase in ice volume can be hindered by external high pressure which suppresses ice nucleation and ice crystal growth. Hence, high pressure is physical cryoprotectant which additionally changes the physical behaviour of water. High pressure lowers both the melting point and the freezing point, with the effect that at 2,048 bar, the melting point of water reaches its minimum of -22 ºC. At the same time, the possibili‐ ty for supercooling is expanded, and the freezing point is shifted to -92 ºC [87]. Consequently, less heat is produced by crystallisation and less heat has to be extracted from the sample during freezing. In practice, it means that only several 1,000 K/s are required to vitrify biological matter at 2,048 bar [85]; thus, the sample thickness that can be vitrified increases tenfold, up to 200 μm [88]. Vitrification process during high-pressure freezing is obtained by pressurising the sample at 2,048 bar and subsequent rapid cooling at -196 ºC using liquid nitrogen jets within tens of milliseconds [72]. However, in very special cases, if the concentration of intrinsic cryoprotec‐ tants is high, samples even as thick as 600 μm can be vitrified [89,90]. Usually, the space around samples during vitrification has to be filled with appropriate filler. The role of the filler is to effectively transfer heat across the sample as well as cryoprotection during cooling process. Therefore, an important issue is physiological compatibility with the cells or tissues [67].

Except that HPF technique enables vitrification of thick samples, this technique has evident limitations. Firstly, gaseous compartments collapse at a high pressure. To counteract this phenomenon, intracellular air-filled spaces (e.g. lung tissue, plant leaves) need to be filled with a chemically inert solvent not mixable with water [91,92]. Secondly, high pressure solidifies the lipid bilayers and leads to a change in biological membrane structure [93]. Similarly, cholesteric liquid crystal of DNA cannot be retained by using HPF [94]. The variety of samples prepared by HPF is enormous, and in the literature, there can be found numerous articles about the advantage of cryoimmobilisation over chemical fixation. Naturally, some exceptions to this rule exist. The HPF technique is less successful for labile tissue, such as brain or nerve tissue, where excision of tissue can last for quite some time, which results in anoxia and osmotic effects [95]. The inner ear tissue is another example of difficult to preserve sample by HPF [96]. In the case of stereocilia, which are mechanosens‐ ing organelles of hair cells, the actin core had a distorted appearance after freezing, described as 'tangled'. Fortunately, a hybrid technique resolves this problem and amplify the strength of HPF as cryopreparation technique (see Sect. 8).

The handling time during sample preparation is crucial because short time enables to obtain sample quality close to the native state. This requirement has brought several interesting solutions for sample preparation step prior HPF. While whole organisms (e.g. *Caenorhabditis elegans*, *Drosophila* [67], zebrafish embryos [97]) or human and animal cell culture is quite easy to load into the carrier, tissue sampling is more demanding. In the former case, cells are either directly loaded into the carriers or first cultured on chosen substrates (e.g. sapphire coverslips [98], Aclar disc [99] or gold grids [100]) and subsequently transferred to the live carriers and vitrified. For different tissues, short handling times can be reduced by using a rapid micro‐ biopsy system. The microbiopsy system allows to accomplish the excision of a small sample and freeze it in less than 30 s [101]. Another solution for improving both the fast sample transfer and freezing reproducibility process is the rapid transfer system [102]. The principal advantage of this tool is the ability to correlate light microscopy and HPF with about 5 s time resolution [103]; thus, a catching of dynamic processes at known point and rare event localisation is possible. For faster cellular events, such as ultrafast endocytosis, flash-and-freeze approach was designed [104]. Flash-and-freeze electron microscopy combines optogenetics with HPF; viz. a brief single light stimulus is applied to the sample with subsequent freezing step. This tool can capture cellular dynamics with millisecond temporal resolution and nanometre spatial resolution.

#### **5.3. Self-pressurised rapid freezing**

addition, the surrounding environment has an influence on the success of the final vitrification. In the case of thin films, usually 100 nm thick, prepared before plunging, sample is subjected to evaporation and heat transfer processes. In consequence, sample is exposed to risk of concentration, pH, ionic strength and temperature changes. Therefore, a new method was devised by Bellare et al. [74] and further improved by Frederik [75], who has constructed a new robot for sample vitrification by this technique. Plunge freezing can be applied to viruses [76], bacteria [77,78], isolated protein complexes [79,80] and whole cells [81,82], however, only 10 μm thick. In the last case, the flat thin periphery of cells [81] or whole thin cells, e.g.

Although high cooling rate values are obtained during vitrification by PF, this technique is suitable for samples as thick as few micrometres. The reason is poor heat conduction properties of water because even an infinitely high cooling rate applied at the sample surface rapidly decays within the sample. Thicker-sample vitrification is only possible when cooling rate is

The high-pressure freezing (HPF) technique was developed in the 1960s by Moor and Riehle [86], but as a routine laboratory technique, it became available from the mid-1980s. This freezing technique is based on Le Chatelier's principle and on the fact that physical properties of water can be manipulated in some range. Water during crystallisation expands its volume and forms crystal. According to Le Chatelier's principle, the increase in ice volume can be hindered by external high pressure which suppresses ice nucleation and ice crystal growth. Hence, high pressure is physical cryoprotectant which additionally changes the physical behaviour of water. High pressure lowers both the melting point and the freezing point, with the effect that at 2,048 bar, the melting point of water reaches its minimum of -22 ºC. At the same time, the possibili‐ ty for supercooling is expanded, and the freezing point is shifted to -92 ºC [87]. Consequently, less heat is produced by crystallisation and less heat has to be extracted from the sample during freezing. In practice, it means that only several 1,000 K/s are required to vitrify biological matter at 2,048 bar [85]; thus, the sample thickness that can be vitrified increases tenfold, up to 200 μm [88]. Vitrification process during high-pressure freezing is obtained by pressurising the sample at 2,048 bar and subsequent rapid cooling at -196 ºC using liquid nitrogen jets within tens of milliseconds [72]. However, in very special cases, if the concentration of intrinsic cryoprotec‐ tants is high, samples even as thick as 600 μm can be vitrified [89,90]. Usually, the space around samples during vitrification has to be filled with appropriate filler. The role of the filler is to effectively transfer heat across the sample as well as cryoprotection during cooling process. Therefore, an important issue is physiological compatibility with the cells or tissues [67].

Except that HPF technique enables vitrification of thick samples, this technique has evident limitations. Firstly, gaseous compartments collapse at a high pressure. To counteract this phenomenon, intracellular air-filled spaces (e.g. lung tissue, plant leaves) need to be filled with a chemically inert solvent not mixable with water [91,92]. Secondly, high pressure solidifies the lipid bilayers and leads to a change in biological membrane structure [93]. Similarly, cholesteric liquid crystal of DNA cannot be retained by using HPF [94]. The

*Ostreococcus* [83] and *Plasmodium* sporozoites [84], is directly observed.

reduced under particular conditions [60,85].

204 The Transmission Electron Microscope – Theory and Applications

**5.2. High-pressure freezing**

Self-pressurised rapid freezing (SPRF) has been introduced in 2007 as a novel and low-cost cryofixation method. Sample is loaded inside the copper tube which is then clamp sealed at both ends and plunged directly into the cryogen [105]. The essence of vitrification by SPRF is isochoric freezing process. This technique is based on the fact that water expands upon freezing, and formed crystal ice builds pressure inside the confined volume of the tube. Thus, forma‐ tion of internal pressure by hexagonal ice and/or supercooled water expansion supports the vitrification of the sample, at least in parts of the tube. Additionally, growing ice crystals close to the tube wall might concentrate the cryoprotectant in the centre and support vitrification [106]. Another factor that influence cryofixation is higher heat diffusion coefficient of crystal‐ line ice, and this means a faster cooling system [107]. Although this explanation is still only hypothetical, the fact is that for vitrification by SPRF, lower concentration of dextran is required in comparison with HPF, thereby causing less osmotic changes [108]. The ultrastructural quality of vitrified specimens by SPRF and HPF is comparable, although on average, only 50% of the structures will be vitrified and 50% will be in crystalline form. To prove the concept of selfpressurised vitrification, different organisms were used including bacterial strains, yeast cells, eukaryotic cell culture and *C. elegans* nematodes. In the case of eukaryotic cell culture, it was noted that the copper tubes are poisonous, even in a typical time range (30–60 s) during sample preparation prior vitrification. Thus, the physiological or 'close-to-native' state is definitely lost [106]. An aluminium container was proposed as a substitute for poisonous copper tubes, although authors suggested developments of new biocompatible coatings of the copper tube's internal surface. The reason was a slightly higher degree of crystalline ice volume in alumini‐ um tubes. The necessity to prepare cell culture as a suspension is a different obstacle; there‐ fore, although SPRF technique is cheaper and less laborious in comparison with HPF, this technique cannot be used in morphological research where cell adhesion is of critical point. More recently, SPRF was used to vitrify *Arabidopsis* inflorescence stems with subsequent cryosectioning in order to obtain architecture information about plant cell walls. This knowl‐ edge is another step to improve understating of biofuel plant material and to rationalise reengineering of second-generation lignocellulosic biofuel crops [109].

### **6. Post-freezing processing**

After vitrification, water within the samples is amorphous as long as samples are stored below devitrification temperature (<-136 ºC). Further processing of the sample depends on its size and aims of the study. The electron beam penetration limit in TEM is about 1 μm. Therefore, many different small-enough samples, such as macromolecular complexes, viruses or bacteria as well as subcellular structures at a periphery of the cell, can be imaged directly under cryo-TEM. Water inside a column of TEM evaporates, unless a special cold stage holder is used at temperature close to that of liquid nitrogen. At this temperature, water evaporation rate is almost negligible and vitrified material can be observed without loss of image quality [110]. However, thicker samples require distinct follow-up procedures – specimen sectioning is the basic need. Vitrified material can be sliced at a temperature below devitrification point into ultrathin sections and subsequently analysed under cryo-microscope (CEMOVIS). As another option, the focused ion beam technology is adequate for the purpose of thinning vitreously frozen biological material. Besides application of pure cryotechniques, specimens can be prepared by cryo-to-room temperature techniques.

#### **6.1. CEMOVIS**

In principle, cryo-electron microscopy of vitreous sections (CEMOVIS) technique is as simple as possible. The first step is vitrification of biological material; then vitrified material is sectioned in a cryo-ultramicrotome. Frozen sections are transferred to a grid, and subsequently the grid is imaged in a cryo-TEM under low-dose condition [111]. Critical issue leading to final success is hidden in performing all these manipulations below devitrification temperature. In spite of the fact that CEMOVIS seems to be a simple technique, the modern state of this approach was strongly connected with indispensable laboratory work and technological development of electron microscopy field.

In the first cryosectioning-based studies [42,43], samples were freeze-dried before analysis under electron microscope. Another important development was sectioning of unfixed biological material at -150 ºC [112], together with transfer and observation of frozen-hydrated specimens in the transmission electron microscope [113]. Similar to the previous studies, also in these works, obtained sections were not analysed as vitreous specimens.

structures will be vitrified and 50% will be in crystalline form. To prove the concept of selfpressurised vitrification, different organisms were used including bacterial strains, yeast cells, eukaryotic cell culture and *C. elegans* nematodes. In the case of eukaryotic cell culture, it was noted that the copper tubes are poisonous, even in a typical time range (30–60 s) during sample preparation prior vitrification. Thus, the physiological or 'close-to-native' state is definitely lost [106]. An aluminium container was proposed as a substitute for poisonous copper tubes, although authors suggested developments of new biocompatible coatings of the copper tube's internal surface. The reason was a slightly higher degree of crystalline ice volume in alumini‐ um tubes. The necessity to prepare cell culture as a suspension is a different obstacle; there‐ fore, although SPRF technique is cheaper and less laborious in comparison with HPF, this technique cannot be used in morphological research where cell adhesion is of critical point. More recently, SPRF was used to vitrify *Arabidopsis* inflorescence stems with subsequent cryosectioning in order to obtain architecture information about plant cell walls. This knowl‐ edge is another step to improve understating of biofuel plant material and to rationalise

After vitrification, water within the samples is amorphous as long as samples are stored below devitrification temperature (<-136 ºC). Further processing of the sample depends on its size and aims of the study. The electron beam penetration limit in TEM is about 1 μm. Therefore, many different small-enough samples, such as macromolecular complexes, viruses or bacteria as well as subcellular structures at a periphery of the cell, can be imaged directly under cryo-TEM. Water inside a column of TEM evaporates, unless a special cold stage holder is used at temperature close to that of liquid nitrogen. At this temperature, water evaporation rate is almost negligible and vitrified material can be observed without loss of image quality [110]. However, thicker samples require distinct follow-up procedures – specimen sectioning is the basic need. Vitrified material can be sliced at a temperature below devitrification point into ultrathin sections and subsequently analysed under cryo-microscope (CEMOVIS). As another option, the focused ion beam technology is adequate for the purpose of thinning vitreously frozen biological material. Besides application of pure cryotechniques, specimens can be

In principle, cryo-electron microscopy of vitreous sections (CEMOVIS) technique is as simple as possible. The first step is vitrification of biological material; then vitrified material is sectioned in a cryo-ultramicrotome. Frozen sections are transferred to a grid, and subsequently the grid is imaged in a cryo-TEM under low-dose condition [111]. Critical issue leading to final success is hidden in performing all these manipulations below devitrification temperature. In spite of the fact that CEMOVIS seems to be a simple technique, the modern state of this approach was strongly connected with indispensable laboratory work and technological

reengineering of second-generation lignocellulosic biofuel crops [109].

**6. Post-freezing processing**

206 The Transmission Electron Microscope – Theory and Applications

prepared by cryo-to-room temperature techniques.

development of electron microscopy field.

**6.1. CEMOVIS**

However, the first successful trails in cryosectioning of vitreous material were built on precursors' work and existing cryo-microtomes and took place in the eighties of the last century [111,114]. Obtained good results were not reproduced each time, and they were difficult to extend to other specimens [70]. It took another 20 years, where new vitrifica‐ tion techniques were established, and scientists obtained better knowledge of the vitreous state of water and cryosectioning method was optimised. Although promising results have been obtained at the beginning of this century, 2004 was a pivotal year because CEMO‐ VIS has reached maturity. Since then, CEMOVIS is regarded as a routine laboratory technique, at least by some researchers, and reveals the native state of cells and tissues with remarkable quality and resolution [115].

Frozen-hydrated sections are prepared without any additional purification, fixation, dehy‐ dration and staining steps. As a result, all cellular components are immobilised inside vitrified water, with preservation to the atomic level [116]. During vitrification process, using HPF or SPRF, usually a 20% solution of a high-molecular-mass dextran is used. Addition of dextran to the sample mimics the vitrification properties of typical cells. Consequently, surrounding environment of the cells is vitrified and thus eliminates possibility of crystalline ice formation. Hexagonal or cubic ices make the sample brittle for sectioning. From another point of view, addition of dextran polymer ensures better sectioning of the sample [117]. In spite of the fact that observed specimen remains fully hydrated, unstained and close to the native state, it is not absolutely free of artefacts.

Vitreous water is a liquid with very high viscosity. This viscous nature of water entails difficulties during cryosectioning process and results in cutting-induced deformation. Resinembedded and vitrified materials differ in cutting properties. During the cutting process of resin-embedded material, obtained thin sections from diamond knife are subsequently straightened during floating on the water surface through a high surface tension [118]. The required liquid for CEMOVIS techniques probably does not exist [117]. Therefore, a dry diamond knife is dedicated for cryosectioning of vitreous samples. The absence of liquid to float vitreous sections results in increasing interaction between the forming section and the knife [118]. These factors results in the formation of cutting artefacts such as chatter, compres‐ sion, knife marks and crevasses [119]. In fact, sectioning of vitreous biological samples is technically difficult, but what is more important is some of the artefacts can be eliminated to some extent [119,120]. For example, knife marks are reduced through the use of undamaged diamond knife and through elimination of frost and debris at the edge of the knife [121]. The humidity inside the chamber of the cryo-ultramicrotome can be reduced to below 1% by using a protective glove box surrounding the cryo-ultramicrotome. It ensures that cryo-tools remain clean from frost contamination during vitreous sectioning [122]. The forming section flows during cutting process, thus applied force yields in different deformations. Crevasses are fractures on the surface of the section and they are more severe in thick sections and high cutting speed [115,119]. To minimise this artefact, sections should be thick but less than 70 nm. Chatter is defined as a periodic variation in section thickness along the cutting direction and depends on the gliding properties of the knife surface. It can be minimised by increasing cutting speed [117,123]. Chatter is associated with irregular compression which is considered as the most prominent cutting artefact. Compression is formed due to irregular friction of the knife surface, and it makes the section shorter along cutting direction compared to the length of sample's block face. The main problem associated with compression is its heterogeneous nature which is discerned at cellular and molecular level [119,124–127]. Microtubule may serve as an example [124,125]: very often microtubules were observed as noncircular-shaped structures in case when long axes of these structures were not perpendicular to the cutting direction. What is more interesting, microtubules that were close to each other did not always display the same degree of deformation. A more detailed explanation can be found in [125]. Compression is reduced either through the use of low-angle knives [128] or through increasing the cutting thickness. Each sample is different, and different combinations of artefacts are possible. For that reason, it is important to find appropriate sectioning conditions to obtain optimal results [126]. Another solution is application of an oscillating knife to reduce cuttinginduced artefacts of vitreous sections [120].

Probably the most difficult step during cryo-ultramicrotomy is transfer of vitreous ribbon of sections onto the carbon grid. To accomplish this step successfully, different approaches have been developed. One of the solutions is micromanipulator to hold and control the vitreous cryosections by eyelash when they come off from the knife edge. During the entire process, the ribbon is under constant tension as it grows longer, thereby keeping the ribbon as straight as possible. When the appropriate length is obtained, the ribbon is attached to the grid surface by lowering the micromanipulator and holding it in optimal position while a second eyelash is used to affix the other end of the ribbon to the grid surface from the knife edge. It is possible to affix a few ribbons to a single grid. However, the discussed solution is time consuming and prone to ice contamination [129]. The ribbons on the grid can be flattened by pressing with tools. The aim of this step is to reduce the probability of losing the sections during storage and transfer the grid, as well as improve the stability of section under electron beam [117]. Another solution is electrostatic charging for attaching the sectioned ribbon to the grid [122]. In comparison with micromanipulator solution, this method increases the successful attachment of frozen-hydrated sections to the carbon film, albeit both methods cannot guarantee uniform attachment of cryosections to the carbon film. This results in higher sensitivity of the section to the beam exposure and section movement during image acquisition, especially during electron tomography [123]. Recently, a new tool based on an aforementioned solution was presented [130]. One of the micromanipulator is used to manipulate the section ribbon by electrically conductive fibre; the second one positions the grid beneath the newly formed ribbon, and with the help of an ioniser, the ribbon is attached to the grid. This tool greatly facilitates manipulations, but sectioning artefacts remain. In summary, although much is known about CEMOVIS procedure and its artefacts, there is no remedy so far for discussed limitations. For this reason, CEMOVIS is still far from a routine application and general use.

#### **6.2. Focused ion beam milling of vitreous samples**

fractures on the surface of the section and they are more severe in thick sections and high cutting speed [115,119]. To minimise this artefact, sections should be thick but less than 70 nm. Chatter is defined as a periodic variation in section thickness along the cutting direction and depends on the gliding properties of the knife surface. It can be minimised by increasing cutting speed [117,123]. Chatter is associated with irregular compression which is considered as the most prominent cutting artefact. Compression is formed due to irregular friction of the knife surface, and it makes the section shorter along cutting direction compared to the length of sample's block face. The main problem associated with compression is its heterogeneous nature which is discerned at cellular and molecular level [119,124–127]. Microtubule may serve as an example [124,125]: very often microtubules were observed as noncircular-shaped structures in case when long axes of these structures were not perpendicular to the cutting direction. What is more interesting, microtubules that were close to each other did not always display the same degree of deformation. A more detailed explanation can be found in [125]. Compression is reduced either through the use of low-angle knives [128] or through increasing the cutting thickness. Each sample is different, and different combinations of artefacts are possible. For that reason, it is important to find appropriate sectioning conditions to obtain optimal results [126]. Another solution is application of an oscillating knife to reduce cutting-

Probably the most difficult step during cryo-ultramicrotomy is transfer of vitreous ribbon of sections onto the carbon grid. To accomplish this step successfully, different approaches have been developed. One of the solutions is micromanipulator to hold and control the vitreous cryosections by eyelash when they come off from the knife edge. During the entire process, the ribbon is under constant tension as it grows longer, thereby keeping the ribbon as straight as possible. When the appropriate length is obtained, the ribbon is attached to the grid surface by lowering the micromanipulator and holding it in optimal position while a second eyelash is used to affix the other end of the ribbon to the grid surface from the knife edge. It is possible to affix a few ribbons to a single grid. However, the discussed solution is time consuming and prone to ice contamination [129]. The ribbons on the grid can be flattened by pressing with tools. The aim of this step is to reduce the probability of losing the sections during storage and transfer the grid, as well as improve the stability of section under electron beam [117]. Another solution is electrostatic charging for attaching the sectioned ribbon to the grid [122]. In comparison with micromanipulator solution, this method increases the successful attachment of frozen-hydrated sections to the carbon film, albeit both methods cannot guarantee uniform attachment of cryosections to the carbon film. This results in higher sensitivity of the section to the beam exposure and section movement during image acquisition, especially during electron tomography [123]. Recently, a new tool based on an aforementioned solution was presented [130]. One of the micromanipulator is used to manipulate the section ribbon by electrically conductive fibre; the second one positions the grid beneath the newly formed ribbon, and with the help of an ioniser, the ribbon is attached to the grid. This tool greatly facilitates manipulations, but sectioning artefacts remain. In summary, although much is known about CEMOVIS procedure and its artefacts, there is no remedy so far for discussed limitations. For this reason, CEMOVIS is still far from a routine application and general use.

induced artefacts of vitreous sections [120].

208 The Transmission Electron Microscope – Theory and Applications

Focused ion beam (FIB) is an alternative method for sample thinning, free of artefact characteristic for CEMOVIS. This technique is widely used in material science; however, Marko et al. [131] proved that FIB milling can be applied in preparing biological materi‐ al. FIB milling of vitreous samples is conducted using a dual-beam microscope. The dualbeam microscope is a combination of FIB system and scanning electron microscope. During FIB milling of vitreous sample, a finely focused beam of ions, usually gallium, is used to ablate the surface of the specimen through sputtering process. The whole procedure is under visual control of the SEM to ensure optimal procedure of sample preparation [132]. However, direct interaction between the ion beam and vitreous material must be taken into consideration due to possible sample damage. The application of a gallium ion beam with current of 10 pA and 30 kV acceleration does not cause sample devitrification [131]. Moreover, interaction of FIB with vitreous sample results in implantation of an ion layer, as thick as 5–20 nm into the FIB milling surface [133]. Indeed, the thickness of implanted gallium layer is almost negligible in a vitreous specimen with a thickness of 100–300 nm. Furthermore, the ion layer is much thinner in comparison with crevasses found in vitreous sections.

Cryo-FIB micromachining is a relatively new technique and remains in its early stages. Nevertheless, few sample preparation strategies have been introduced in the last 10 years. To date, bacteria and small eukaryotic cells, like *Mycobacterium smegmatis, Saccharomyces cerevi‐ siae* and *Dictyostelium discoideum* [134,135], *Escherichia coli*, HeLa cells [136,137], BHK-21 cells [138] and *Aspergillus niger* [139], are deposited for culturing on the TEM grids and vitrified by plunge freezing technique. Next, vitrified material is transferred into the dual-beam micro‐ scope for a thinning process with a precision in the 10–100 nm range. At this stage, different FIB milling strategies for vitrified cellular samples are possible. The optimal geometry for small prokaryotic cells is wedge shaped, where frozen material is sputtered away at an oblique angle (~10º) with respect to the plane of the grid. Consequently, wedge-shaped vitreous material can be imaged with up to 3 μm transparency length and a thickness gradient less than 400 nm [134]. Eukaryotic and other cells similar in shape and size are milled in a thin self-supported membrane. During this process, a specific region of interest is localised and then rectangular sector below and above the selected volume is sputtered away, leaving behind a thin mem‐ brane, commonly referred to as a lamella, supported by the surrounding unmilled cells and ice [135]. Another option to obtain lamellas is a traditional FIB lift-out method, although it was deemed impossible because of the difficulty in obtaining platinum deposition at cryogenic temperatures [139,140]. Shortly, after vitrification, the feature of interest is defined through SEM; next the sample is cryo-coated with platinum (Pt) and two trenches are milled on each side of the lamella to be extracted. In the next steps, the sides and bottom of the lamella are sputtered away, and by using cold nanomanipulator, lamella is then lifted out from the sample and finally attached to the TEM grid by cryo-Pt deposition. During the last step, the attached lamella is thinned enough to be transparent to electron beam. The lamella-based sample preparation has an advantage over the wedge-shaped strategy, because the simple ablation geometry would not permit to easily find and target structures of interests embedded deeply in cellular volume [135,136]. Thick samples and suspension of cells, e.g. *S. cerrevisiae* [140,141] and muscle tissue [142,143], are vitrified in copper tube or dedicated carriers via high-pressure freezing. After that, material hidden inside both the tube and the carrier is exposed during pretrimming step inside the cryo-ultramicrotome. Subsequently, vitreous sample is transfer‐ red to the dual-beam microscope and milled according to H-bar strategy, ultimately resulting in lamellas with the required thickness and surface area. Finally, prepared sample is transfer‐ red to the cryo-TEM in order to perform visualisation under low-dose mode. Critical point is when each of the described steps must be carried under devitrification temperature and minimising frost or warming during transfer steps implicates customised transfer device introduction. For samples vitrified by plunge freezing, different cryo-FIB transfer stations and cryo-FIB shuttles were introduced [135–137], while high-pressure frozen samples require sophisticated and complex transfer systems like cryo-nano-bench system [140] or an inter‐ mediate specimen holder [142,143].

The cryo-FIB technique provides controlled access to specific supramolecular structures buried inside the cell. Moreover, many macromolecular complexes that are present in low copy numbers may be studied in their native cellular environment because homogenously thick lamellas with more than 100 μm2 areas can be prepared in a controlled and targeted manner. However, this preparation technique is low throughput due to several factors. As it was mentioned, an incidence of the ion beam should be as low as possible because a smaller milling angle produces a larger viewing area for analysis and minimises the deposition effect of milled material. The latter factor is not interrelated with milling currents or other parameters during sample thinning as such [135,137]. Inhomogeneous and varied composition of the vitreous samples is the main cause for curtaining effect which finally results in strong inhomogeneous lamella thickness [134,142]. These effect is reduced by deposition of organometallic platinum with a gas injection system without electron or ion beam radiation, prior to lamella prepara‐ tion [144]. Another issue is the amount of information obtained from prepared samples during cryo-electron tomography. During cryo-FIB milling process, part of the vitreous specimen is physically destroyed along all axes [134]. In contrary, after vitreous sectioning, a series of cryosections is obtained, and information along these axes is partly remained, especially along *z*-axis. A further problem is the time needed to prepare sample prior to visualisation. Due to the large size of eukaryotic cells, longer milling time is required in comparison with the prokaryotic samples [135–137]. Other challenges related to the increased size of eukaryotic cells are identifying and targeting specific sites for processing. Small organisms, such as *E. coli*, are readily distinguished from the vitreous ice. Inversely, eukaryotic cells are surrounded by thick ice, thus identifying the area of interest is not simple. A method to overcome this limitation is either milling of many adjacent places to find features of interest or application of correlative light and electron microscopy techniques. Cells are cultured on EM finder grids and optical images are recorded before vitrification. Appropriate regions are selected for subsequent cryo-FIB milling based on light microscopy photos [136,145]. Alternatively, the frozen-hydrated sample is imaged under cryo-fluorescence microscopy before sample preparation by cryo-FIB for further analysis. The second approach allows direct correlation of the prepared vitreous sample between the two imaging modalities [134,138,146]. Moreover, localisation of smaller molecules or structures that exist in low copy number is simplified because targeting is based on clonable labels, such as green fluorescent protein [134]. The new idea presented recently is to localise structure of interest by fluorescent labels using cryo-light microscopy and then use it for coordinate transformation-based approach in the FIB-SEM system for milling [138].

At present, cryo-FIB milling of vitreous samples remains cumbersome and far from routine [143]. Much effort is required to improve the efficiency and repeatability of cryo-FIB milling process, such as reproducibility of lamella quality, i.e. overall thickness and thickness homogeneity, localisation feature of interests or better avoidance of frost during cryotransfer steps. Therefore, the main issue in the presented articles is improving sample preparation protocols. Nevertheless, some interesting results have been achieved. For example, ten nuclear pore complexes in *D. discoideum* cells were identified in situ in one tomogram. Subtomogram averaging process yields the structure of nuclear pore with resolution of 7.9 nm. To achieve 6 nm resolution, 4,182 protomers are required from isolated nuclei using the same type of analysis [135]. Another result comes from bacterial cells, where membrane invaginations into both the cytoplasmic and periplasmic spaces of *E. coli* were found [136]. This technique will greatly facilitate high-resolution imaging of dynam‐ ic process, such as HIV particles travelling into the deep side of the host cell at different stages of infection, especially when viral capsid interacts with nuclear pore complex components [136]. By using FIB milling process, ryanodine receptors in toadfish swimblad‐ der muscle were determined. Obtained results agreed remarkably well with those descri‐ bed previously, albeit further study will be required to understand structural features of ryanodine receptor connected to the T-tubule [143]. Cryo-FIB is also a promising thinning tool for describing new bacterial cytoplasmic structures termed as a stack. Stacks were defined as piles of oval disc subunits which are surrounded by a membrane-like struc‐ ture. These structures are localised in the cell cytoplasm and are presented separately or grouped together in variable number within each cell. One can only speculate about stacks' function, but it is suggested that they could be related to the bacterial cell replication process. Due to compression created during CEMOVIS and visualisation limitations arising from plunged frozen samples, cryo-FIB technique could provide new insight into macromo‐ lecular assembly of membrane-enclosed discs [147].

#### **6.3. Nature of vitreous material**

in cellular volume [135,136]. Thick samples and suspension of cells, e.g. *S. cerrevisiae* [140,141] and muscle tissue [142,143], are vitrified in copper tube or dedicated carriers via high-pressure freezing. After that, material hidden inside both the tube and the carrier is exposed during pretrimming step inside the cryo-ultramicrotome. Subsequently, vitreous sample is transfer‐ red to the dual-beam microscope and milled according to H-bar strategy, ultimately resulting in lamellas with the required thickness and surface area. Finally, prepared sample is transfer‐ red to the cryo-TEM in order to perform visualisation under low-dose mode. Critical point is when each of the described steps must be carried under devitrification temperature and minimising frost or warming during transfer steps implicates customised transfer device introduction. For samples vitrified by plunge freezing, different cryo-FIB transfer stations and cryo-FIB shuttles were introduced [135–137], while high-pressure frozen samples require sophisticated and complex transfer systems like cryo-nano-bench system [140] or an inter‐

The cryo-FIB technique provides controlled access to specific supramolecular structures buried inside the cell. Moreover, many macromolecular complexes that are present in low copy numbers may be studied in their native cellular environment because homogenously thick

However, this preparation technique is low throughput due to several factors. As it was mentioned, an incidence of the ion beam should be as low as possible because a smaller milling angle produces a larger viewing area for analysis and minimises the deposition effect of milled material. The latter factor is not interrelated with milling currents or other parameters during sample thinning as such [135,137]. Inhomogeneous and varied composition of the vitreous samples is the main cause for curtaining effect which finally results in strong inhomogeneous lamella thickness [134,142]. These effect is reduced by deposition of organometallic platinum with a gas injection system without electron or ion beam radiation, prior to lamella prepara‐ tion [144]. Another issue is the amount of information obtained from prepared samples during cryo-electron tomography. During cryo-FIB milling process, part of the vitreous specimen is physically destroyed along all axes [134]. In contrary, after vitreous sectioning, a series of cryosections is obtained, and information along these axes is partly remained, especially along *z*-axis. A further problem is the time needed to prepare sample prior to visualisation. Due to the large size of eukaryotic cells, longer milling time is required in comparison with the prokaryotic samples [135–137]. Other challenges related to the increased size of eukaryotic cells are identifying and targeting specific sites for processing. Small organisms, such as *E. coli*, are readily distinguished from the vitreous ice. Inversely, eukaryotic cells are surrounded by thick ice, thus identifying the area of interest is not simple. A method to overcome this limitation is either milling of many adjacent places to find features of interest or application of correlative light and electron microscopy techniques. Cells are cultured on EM finder grids and optical images are recorded before vitrification. Appropriate regions are selected for subsequent cryo-FIB milling based on light microscopy photos [136,145]. Alternatively, the frozen-hydrated sample is imaged under cryo-fluorescence microscopy before sample preparation by cryo-FIB for further analysis. The second approach allows direct correlation of the prepared vitreous sample between the two imaging modalities [134,138,146]. Moreover, localisation of smaller molecules or structures that exist in low copy number is simplified because targeting is based

areas can be prepared in a controlled and targeted manner.

mediate specimen holder [142,143].

210 The Transmission Electron Microscope – Theory and Applications

lamellas with more than 100 μm2

The sample in native state is very different from what has been seen before with conventional microscopy. Vitrified material is as close as possible to the native state because during sample preparation, neither chemical fixation, staining nor dehydration is used. Therefore, the final images represent the real distribution of the immobilised biological material within vitrified water. With frozen-hydrated samples, the contrast is proportional to the density and distri‐ bution of molecular inhabitants within the thickness of the sections. Moreover, structures in vitreous material are equally visible over the entire thickness, thus the native-state inherent low contrast due to low signal-to-noise ratio. In contrary to stained material, imaging of native biological material relies on phase contrast, which strongly depends on focus [71]. An addi‐ tional issue is plethora of overlapping information for the reason that the typical fine details are much smaller than the section thickness. The solution overcoming this limitation is both preparation of thin sections or lamellas and electron tomography to obtain a three-dimensional model of the material distribution in the vitreous material. Frozen-hydrated specimens behave differently under electron beam in comparison with plastic sections. A characteristic phenom‐ enon that may be developed is bubbling. Bubbling is a result of gas accumulation produced by electron beam decomposition of biological matter. What is the most interesting is that different substances have different electron radioresistance. Another problem is beam-induced deformations which are seen twofold. Vitreous material can be considered as high-viscosity liquid and can be rearranged by the electrons. On the one hand, sharp irregularities, such as crevasses and knife marks, under electron beam are removed from the vitreous section due to increasing the flow of the section. On the other hand, some biological structures, i.e. chromatin, aggregate under the effect of beam producing locally apparent higher-density area [148].

The main disadvantage of frozen-hydrated specimens is its uselessness to perform postimmunolocalisation of studied target. Antibodies require proper conditions for working, that is, ambient conditions and water solution. Accordingly, some researchers have developed a specific label for the identification of molecules for cryo-electron microscopy. These labelling techniques are base either on a clonable ferritin FtnA protein [149] or a biotin-streptavidin complex [150], called STAMP approach [151] and SNAP-tag technology [152]. However, ligand labelling for cryo-EM is still an emerging field; hence, another preparation technique is dedicated for immunogold labelling and structural studies, namely, freeze substitution. Indeed, vitreous and freeze-substituted materials are very complementary [123]. The latter preparation solution should be considered as the method of choice when high-resolution study is not the major aim. Resin-embedded material is easier to obtain and is less sensitive to electron beam in comparison with vitreous material, and for this reason, analysis of larger sample area is possible. Additionally, plastic sections can be thicker than vitreous ones, and thus, the former enable studying a larger volume. The other advantage of resin sections is the possibility of immunogold labelling.

#### **6.4. Freeze substitution**

Freeze substitution (FS) is a hybrid method that bridges the gap between vitrified material and room temperature ultramicrotomy of resin-embedded material. Biological material after vitrification process is gradually dehydrated in the presence of chemical fixatives at low temperature. Later, the whole process is gradually warmed, and finally, the sample is embed‐ ded in resin. This technique was first introduced in 1941, as a preparative technique for light microscopy samples [153]. The potential of FS at electron microscopy filed was explored by Fernandez-Moran [154] and was further developed by others [155–158].

Freeze substitution process consists of dehydration and chemical fixation step followed by either low- or room-temperature embedding in chosen resin. Dehydration process usually starts at -90 ºC in an organic substitution medium containing chemical fixatives. The key point is that sample must be kept below devitrification temperature. In case of biological material, this temperature remains well below -100 ºC. What happens during water substitution at low temperatures is not fully understood. Nevertheless, vitrified water turns into cubic ice, and then transition takes place into hexagonal ice. Cubic ice is a metastable state, thermodynami‐ cally more stable form of water. The second transition process, from cubic to hexagonal form, occurs at around -80 ºC. The most important event during these transitions is that water molecules probably rearrange by rotation with only small transitional displacement which leads to embedding the biological structure by ice without any segregation. Then the result is that the structural preservation is excellent down to molecular dimension. In reality, FS process deals with cubic and sometimes with hexagonal ice but never with vitreous water [60]. This would imply that cubic ice has no significant influence on the observed morphology at the level of resolution of biological samples during FS process under controlled conditions. Besides these theoretical bases, other aspects must be considered.

preparation of thin sections or lamellas and electron tomography to obtain a three-dimensional model of the material distribution in the vitreous material. Frozen-hydrated specimens behave differently under electron beam in comparison with plastic sections. A characteristic phenom‐ enon that may be developed is bubbling. Bubbling is a result of gas accumulation produced by electron beam decomposition of biological matter. What is the most interesting is that different substances have different electron radioresistance. Another problem is beam-induced deformations which are seen twofold. Vitreous material can be considered as high-viscosity liquid and can be rearranged by the electrons. On the one hand, sharp irregularities, such as crevasses and knife marks, under electron beam are removed from the vitreous section due to increasing the flow of the section. On the other hand, some biological structures, i.e. chromatin, aggregate under the effect of beam producing locally apparent higher-density area [148].

212 The Transmission Electron Microscope – Theory and Applications

The main disadvantage of frozen-hydrated specimens is its uselessness to perform postimmunolocalisation of studied target. Antibodies require proper conditions for working, that is, ambient conditions and water solution. Accordingly, some researchers have developed a specific label for the identification of molecules for cryo-electron microscopy. These labelling techniques are base either on a clonable ferritin FtnA protein [149] or a biotin-streptavidin complex [150], called STAMP approach [151] and SNAP-tag technology [152]. However, ligand labelling for cryo-EM is still an emerging field; hence, another preparation technique is dedicated for immunogold labelling and structural studies, namely, freeze substitution. Indeed, vitreous and freeze-substituted materials are very complementary [123]. The latter preparation solution should be considered as the method of choice when high-resolution study is not the major aim. Resin-embedded material is easier to obtain and is less sensitive to electron beam in comparison with vitreous material, and for this reason, analysis of larger sample area is possible. Additionally, plastic sections can be thicker than vitreous ones, and thus, the former enable studying a larger volume. The other advantage of resin sections is the possibility of

Freeze substitution (FS) is a hybrid method that bridges the gap between vitrified material and room temperature ultramicrotomy of resin-embedded material. Biological material after vitrification process is gradually dehydrated in the presence of chemical fixatives at low temperature. Later, the whole process is gradually warmed, and finally, the sample is embed‐ ded in resin. This technique was first introduced in 1941, as a preparative technique for light microscopy samples [153]. The potential of FS at electron microscopy filed was explored by

Freeze substitution process consists of dehydration and chemical fixation step followed by either low- or room-temperature embedding in chosen resin. Dehydration process usually starts at -90 ºC in an organic substitution medium containing chemical fixatives. The key point is that sample must be kept below devitrification temperature. In case of biological material, this temperature remains well below -100 ºC. What happens during water substitution at low temperatures is not fully understood. Nevertheless, vitrified water turns into cubic ice, and then transition takes place into hexagonal ice. Cubic ice is a metastable state, thermodynami‐

Fernandez-Moran [154] and was further developed by others [155–158].

immunogold labelling.

**6.4. Freeze substitution**

From the biophysical side, low temperature influences on ultrastructure preservation through hydration shell preservation and infiltration of chemical fixatives. FS process preserves the hydration shell at least partially, although less hydrophilic organic solvents are used, e.g. acetone or methanol. It is well known that organic solvents cause protein aggregation, and chemical fixatives react relatively slowly; therefore, it cannot preserve all the cellular components simultaneously. The consequences of chemical fixation are seen as osmotic changes and redistribution or extraction different molecules, i.e. lipids and ions [85,157,159,160]. The reason for the superior structural preservation is infiltration of stabilising or fixative compounds together with the dehydrating agents. After raising the temperature, fixatives react in situ, between -90 ºC and -30 ºC, avoiding penetration and diffusion artefacts [157]. Many different substitution media compositions were developed. The main fixatives are GA, osmium tetroxide and UA used in different combinations and at different percentages in acetone, methanol or ethanol [67,161–164]. Acetone is the most commonly used dehydrating agent because it substitutes at a slower rate than methanol, thus resulting in better structural preservation [165, 166]. The most-used fixatives are OsO4 with or without UA in acetone for morphological study and low concentrated GA in acetone for immunolabelling detection [167–169]. The interest‐ ing observation concerning reactivity of used fixatives was made. Osmium tetroxide at low temperature does not react as protease, but begins cross-linking, via cis-diol covalent bond‐ ing to unsaturated lipid chains at -70 ºC [170]. Uranyl acetate binds to proteins and phospholi‐ pids at an even lower temperature. Glutaraldehyde starts cross-linking at -50 ºC, but it acts readily only at or above -30 ºC [161]. For some immunolabelling study, pure solvent pre‐ serves well both the antigenicity and the ultrastructure of the cells [159,171]. A distinct feature of well-vitrified samples, without visible ice segregation artefact, is a 'reverse contrast' seen as pale membrane against a more electron dense background [172,173]. In order to improve the membrane contrast and preservation of cultured cells, a few different substitution media were introduced. The most interesting substitution medium contains 5% water in acetone [173]. It has been proved that substitution cocktails can include up to 20% of water without deleteri‐ ous ice damage or extending dehydration time [174]. These protocols totally contradict the longstanding theory that a low amount of water as 1% in acetone extends freeze-substitution process four times [161]. A disadvantage of a medium containing water is antigen loss by extraction. Similar results were observed for pure solvents; thus, low concentration of glutaraldehyde can be generally used [175], but the exact substitution protocol requires an individual approach. Alternatively, membrane contrast can be enhanced by using tannic acid-mediated osmium impregnation method [176], tannic acid in acetone during FS [95,177] or different combina‐

tions of glutaraldehyde, UA and OsO4 [178,179]. An attractive protocol omitting osmium tetroxide is based on 20% Araldite/Epon in acetone as fixative in the first step and subsequent embedding in pure epoxide resin. Epoxide compounds react with proteins and lipids and provide interesting results which may become an important tool in getting information about influence of different reagents and protocols on ultrastructure preservation [180]. Current FS procedures are measured in wide time range from less than 24 h for cell culture [162,172,174,181] to longer period of time such as four days for plant material [182,183]. However, other tissues are usually substituted during 2–4 days, e.g. rat liver [184], mouse skin [185] or *C. elegans* [186]. On the one hand, such a wide range of different fixative cocktails gives opportunity to study different structures of interests in both structural and immunolocalisation research. On the other hand, the variety of possibilities become a challenging task for optimisation of FS process because each sample is unique [67].

The last step during sample preparation for the room-temperature ultramicrotomy is sample embedding. Epoxide-based polymers are dedicated for morphological analysis including electron tomography by virtue of a larger stability in the electron beam and ease of section‐ ing. In contrast to epoxide resins, methacrylates do not bind covalently to cellular structures; hence, the antigens of interests remain unaltered and section surface has higher roughness, thereby higher access to antigens. Another advantage of methacrylic resins is low-tempera‐ ture embedding and polymerisation by UV light (Figures 2 and 3); thus, harmful heat effects on epitopes is avoided [157,165]. However, this division is not a rule, because Epon sections were used to identify the subcellular localisation of proteins [187], lipids [188] or carbohy‐ drates [189]. Special attention should be directed to McDonald's FS protocols [168,190]. Super-Quick FS takes only about 6 h from freezing process to resin blocks preparation ready to section. Substitution process from -90 ºC to 0 ºC is performed during 2,5 h; then rapid infiltration in LR White or Epon resin takes another two hours followed by polymerisation at 100 ºC for 1,5–2 h. For this protocol, organisms considered as difficult to fix were chosen. As a result, presented ultrastructure preservation was comparable to standard FS protocols, and high-temperature polymerisation does not affect antigen preservation.

### **7. Sample examples prepared by cryotechniques**

High-resolution study requires superior sample preparation via vitrification process. Sample such as protein, protein complexes, viruses, bacteria or organelles in vitro and in situ within whole organisms or single cell are prepared by plunge freezing. Moreover, in modern structural biology, the main goal is in situ structure determination within unperturbed cells because purified objects are disintegrated during sample purification. Thus, for structural biology at cellular and tissue level, thicker samples are prepared by CEMOVIS (30–200 nm) or cryo-FIB (100–300 nm) techniques. The visualisation of frozen-hydrated biological samples is performed by single-particle analysis or cryo-electron tomography. The former technique enables to achieve near-atomic resolution and is applied to purified viruses, macromolecular complexes and single proteins [80,191,192]. On the other hand, cryo-ET bridges the gap between cellular ultrastructure and the structural analysis of macromolecular complexes within the cell with resolution in the sub-nanometre [193] to 10 nm range [194–196]. Electron tomography of plastic sections is another technique dedicated for cellular structural biology, where the more important aim is to reveal functional-morphological relationships than macromolecular details. Sample is prepared by cryoimmobilisation followed by freezesubstitution process. Next, polymerised specimen is sectioned in 100–400 nm range and analysed [194]. Combination of subsequent section tomograms extended the depth of analysed volume to several micrometres. These advantages led to large-scale imaging where both detail and overview are necessary. Another worth-noting point for ET study is a scanning transmis‐ sion electron microscopy (STEM) mode. Due to lack of inelastic scattering of electrons and chromatic aberration in a STEM mode, resin sections as thick as 1 μm can be analysed [197– 200]. Besides three-dimensional analysis, both frozen-hydrated and plastic sections are also analysed at lower resolution at 2D morphological level.

tions of glutaraldehyde, UA and OsO4 [178,179]. An attractive protocol omitting osmium tetroxide is based on 20% Araldite/Epon in acetone as fixative in the first step and subsequent embedding in pure epoxide resin. Epoxide compounds react with proteins and lipids and provide interesting results which may become an important tool in getting information about influence of different reagents and protocols on ultrastructure preservation [180]. Current FS procedures are measured in wide time range from less than 24 h for cell culture [162,172,174,181] to longer period of time such as four days for plant material [182,183]. However, other tissues are usually substituted during 2–4 days, e.g. rat liver [184], mouse skin [185] or *C. elegans* [186]. On the one hand, such a wide range of different fixative cocktails gives opportunity to study different structures of interests in both structural and immunolocalisation research. On the other hand, the variety of possibilities become a challenging task for optimisation of FS process because

The last step during sample preparation for the room-temperature ultramicrotomy is sample embedding. Epoxide-based polymers are dedicated for morphological analysis including electron tomography by virtue of a larger stability in the electron beam and ease of section‐ ing. In contrast to epoxide resins, methacrylates do not bind covalently to cellular structures; hence, the antigens of interests remain unaltered and section surface has higher roughness, thereby higher access to antigens. Another advantage of methacrylic resins is low-tempera‐ ture embedding and polymerisation by UV light (Figures 2 and 3); thus, harmful heat effects on epitopes is avoided [157,165]. However, this division is not a rule, because Epon sections were used to identify the subcellular localisation of proteins [187], lipids [188] or carbohy‐ drates [189]. Special attention should be directed to McDonald's FS protocols [168,190]. Super-Quick FS takes only about 6 h from freezing process to resin blocks preparation ready to section. Substitution process from -90 ºC to 0 ºC is performed during 2,5 h; then rapid infiltration in LR White or Epon resin takes another two hours followed by polymerisation at 100 ºC for 1,5–2 h. For this protocol, organisms considered as difficult to fix were chosen. As a result, presented ultrastructure preservation was comparable to standard FS protocols, and high-temperature

High-resolution study requires superior sample preparation via vitrification process. Sample such as protein, protein complexes, viruses, bacteria or organelles in vitro and in situ within whole organisms or single cell are prepared by plunge freezing. Moreover, in modern structural biology, the main goal is in situ structure determination within unperturbed cells because purified objects are disintegrated during sample purification. Thus, for structural biology at cellular and tissue level, thicker samples are prepared by CEMOVIS (30–200 nm) or cryo-FIB (100–300 nm) techniques. The visualisation of frozen-hydrated biological samples is performed by single-particle analysis or cryo-electron tomography. The former technique enables to achieve near-atomic resolution and is applied to purified viruses, macromolecular complexes and single proteins [80,191,192]. On the other hand, cryo-ET bridges the gap between cellular ultrastructure and the structural analysis of macromolecular complexes

each sample is unique [67].

214 The Transmission Electron Microscope – Theory and Applications

polymerisation does not affect antigen preservation.

**7. Sample examples prepared by cryotechniques**

CEMOVIS has already provided unusual views of different structures with a molecular resolution in native cellular context including microtubules, mitochondria, Golgi apparatus [194] or desmosomes [116]. Combination of CEMOVIS with electron tomography, called TOVIS [201], becomes a powerful tool in virology [76], microbiology, cellular biology and tissues. The reader is referred to review articles that cover mentioned topics with further references [79,202–205] due to limited space. However, two examples of CEMOVIS application will be pointed out. Cryo-EM of vitreous sections gives opportunity to study different tissues, including skin biopsies. The unravelling of molecular organisation of the skin lipids will significantly improve molecular understanding of the tissue. Until recently, six theoretical models for the molecular arrangement of the extracellular lipid matrix have been proposed. Nevertheless, combination of TOVIS, molecular modelling and EM simulations has revealed a new model of lipid organisation, which rationalises the skin nature and functions [206]. These results will influence dermatology field and thereby further translates in technological developments of new, transdermal drug delivery systems, the development of noninvasive diagnostic sensors and dealing with toxicity from topical exposure to chemicals. Without CEMOVIS, it would be impossible to perform the simultaneous quantification of water and elements in the native state [63]. In combination with fluorescence microscopy, it was showed that the induction of nucleolar stress in cancer cells resulted in both an increase in water content and a decrease in the element content in all cell compartments. The presented study opens new way to understand cell functions, and future research could extend our knowledge about cell activities, depending on actual concentration of ions and the hydration status.

The classical examples of superiority of the cryopreparation techniques, based on HPF followed by FS process, over conventional TEM were showed on the method-dependent bacterial mesosomes [55] and articular cartilage [85], although in the former, further investi‐ gation led to revisiting the mesosome as a site of hydrogen accumulation [208] using quickfreezing preparation of TEM. Among the advantages of plastic sections, one is the section thickness; thus, comparatively large cellular volume can be analysed that was exemplified by microtubule cytoskeleton architecture in yeast [209], organellar relationship in the Golgi region of the pancreatic cells [210] or architecture of the caveolar system [211]. Besides, the reader can find many articles where either different techniques were compared [55,91,96,212–216] or examples were collected in reviews [194,217].

mesosome as

level.

necessary. Another worth-noting point for ET study is a scanning transmission electron microscopy (STEM) mode. Due to lack of inelastic scattering of electrons and chromatic aberration in a STEM mode, resin sections as thick as 1 µm can be analysed [197–200]. Besides three-dimensional analysis, both frozen-hydrated and plastic sections are also analysed at lower resolution at 2D morphological

CEMOVIS has already provided unusual views of different structures with a molecular resolution in native cellular context including microtubules, mitochondria, Golgi apparatus [194] or desmosomes [116]. Combination of CEMOVIS with electron tomography, called TOVIS [201], becomes a powerful tool in virology [76], microbiology, cellular biology and tissues. The reader is referred to review articles that cover mentioned topics with further references [79,202–205] due to limited space. However, two examples of CEMOVIS application will be pointed out. Cryo-EM of vitreous sections gives opportunity to study different tissues, including skin biopsies. The unravelling of molecular organisation of the skin lipids will significantly improve molecular understanding of the tissue. Until recently, six theoretical models for the molecular arrangement of the extracellular lipid matrix have been proposed. Nevertheless, combination of TOVIS, molecular modelling and EM simulations has revealed a new model of lipid organisation, which rationalises the skin nature and functions [206]. These results will influence dermatology field and thereby further translates in technological developments of new, transdermal drug delivery systems, the development of noninvasive diagnostic sensors and dealing with toxicity from topical exposure to chemicals. Without CEMOVIS, it would be impossible to perform the simultaneous quantification of water and elements in the native state [63]. In combination with fluorescence microscopy, it was showed that the induction of nucleolar stress in cancer cells resulted in both an increase in water content and a decrease in the element content in all cell compartments. The presented study opens new way to understand cell functions,

The classical examples of superiority of the cryopreparation techniques, based on HPF followed by FS process, over conventional TEM were showed on the method-dependent bacterial mesosomes [55] and articular cartilage [85], although in the former, further investigation led to revisiting the

**Figure 2.** Ultrastructural demonstration of immunogold labelling of visfatin particles or small clusters consisting of number gold particles were demonstrated in the subcellular compartments of human colorectal HCT-116 mono‐ nucleated cells which were cultured in log phase of growth (Cy cytoplasm, N nucleus, Scale bars: a) 1 μm, b) 500 nm). Cells were chemically fixed, dehydrated and embedded in LR White resin [207]. Visfatin is an enzyme which overex‐ pression is correlated with poor prognosis in cancer patients. In this study, we tried to explore the association between visfatin distribution in subcellular compartments and increased apoptosis in cells treated with cytochalasin B. For fur‐ ther study, our aim is to optimise of HPF-FS protocols for cytoskeleton ultrastructure. Figure and legend adapted and changed from [207] under terms of the CC BY 3.0 license. clusters consisting of number gold particles were demonstrated in the subcellular compartments of human colorectal HCT-116 mononucleated cells which were cultured in log phase of growth (Cy cytoplasm, N nucleus, Scale bars: a) 1 μm, b) 500 nm). Cells were chemically fixed, dehydrated and embedded in LR White resin [207]. Visfatin is an enzyme which overexpression is correlated with poor prognosis in cancer patients. In this study, we tried to explore the association between visfatin distribution in subcellular compartments and increased apoptosis in cells treated with cytochalasin B. For further study, our aim is to optimise of HPF-FS protocols for cytoskeleton ultrastructure. Figure and legend adapted and changed from [207] under terms of the CC BY 3.0 license

**Figure 2a, b** Ultrastructural demonstration of immunogold labelling of visfatin particles or small

 **Figure 3a, b** Human colorectal HCT-116 were vitrified by HPF followed by FS and embedding in LR White resin. During FS process, cells were fixed with only 0,25 % GA. In comparison with cells **Figure 3.** Human colorectal HCT-116 were vitrified by HPF followed by FS and embedding in LR White resin. During FS process, cells were fixed with only 0,25 % GA. In comparison with cells prepared by conventional method (Figure 2), vitrified and freeze-substituted material show outstanding ultrastructure preservation. Even without OsO4 fixation, membranes are clearly visible, mainly rough endoplasmic reticulum and nuclear membranes. Nevertheless, further protocol optimisation is required.

prepared by conventional method (Figure 2), vitrified and freeze-substituted material show outstanding ultrastructure preservation. Even without OsO4 fixation, membranes are clearly visible, mainly rough endoplasmic reticulum and nuclear membranes. Nevertheless, further protocol

a site of hydrogen accumulation [208] using quick-freezing preparation of TEM. Among the

#### optimisation is required **8. Unity is strength: The hybrid techniques**

advantages of plastic sections, one is the section thickness; thus, comparatively large cellular volume can be analysed that was exemplified by microtubule cytoskeleton architecture in yeast [209], organellar relationship in the Golgi region of the pancreatic cells [210] or architecture of the caveolar system [211]. Besides, the reader can find many articles where either different techniques were Immunoelectron microscopy bridges the information gap between molecular biology and ultrastructural studies providing information regarding the function of the internal structures

compared [55,91,96,212–216] or examples were collected in reviews [194,217].

of the cell. The main requirements to obtain valuable results are suitable fixation protocol and functional antibody conjugated with an appropriate nanoparticle to be detected. For many years different protocols were developed for structure localisation study [167] based mainly on Tokuyasu technique, HPF followed by FS and low-temperature embedding [169] or conventional chemical fixation [207,218] together with progressive-lowering temperature technique. Nevertheless, this field is still amenable to new solutions, especially for difficultto-fix samples and difficult-to-fix antigens. The hybrid techniques combine advantages of different cryopreparation techniques in order to eliminate the particular step limitations of each one, at least in part.

necessary. Another worth-noting point for ET study is a scanning transmission electron microscopy (STEM) mode. Due to lack of inelastic scattering of electrons and chromatic aberration in a STEM mode, resin sections as thick as 1 µm can be analysed [197–200]. Besides three-dimensional analysis, both frozen-hydrated and plastic sections are also analysed at lower resolution at 2D morphological

CEMOVIS has already provided unusual views of different structures with a molecular resolution in native cellular context including microtubules, mitochondria, Golgi apparatus [194] or desmosomes [116]. Combination of CEMOVIS with electron tomography, called TOVIS [201], becomes a powerful tool in virology [76], microbiology, cellular biology and tissues. The reader is referred to review articles that cover mentioned topics with further references [79,202–205] due to limited space. However, two examples of CEMOVIS application will be pointed out. Cryo-EM of vitreous sections gives opportunity to study different tissues, including skin biopsies. The unravelling of molecular organisation of the skin lipids will significantly improve molecular understanding of the tissue. Until recently, six theoretical models for the molecular arrangement of the extracellular lipid matrix have been proposed. Nevertheless, combination of TOVIS, molecular modelling and EM simulations has revealed a new model of lipid organisation, which rationalises the skin nature and functions [206]. These results will influence dermatology field and thereby further translates in technological developments of new, transdermal drug delivery systems, the development of noninvasive diagnostic sensors and dealing with toxicity from topical exposure to chemicals. Without CEMOVIS, it would be impossible to perform the simultaneous quantification of water and elements in the native state [63]. In combination with fluorescence microscopy, it was showed that the induction of nucleolar stress in cancer cells resulted in both an increase in water content and a decrease in the element content in all cell compartments. The presented study opens new way to understand cell functions, and future research could extend our knowledge about cell activities, depending on actual

The classical examples of superiority of the cryopreparation techniques, based on HPF followed by FS process, over conventional TEM were showed on the method-dependent bacterial mesosomes [55] and articular cartilage [85], although in the former, further investigation led to revisiting the

**Figure 2.** Ultrastructural demonstration of immunogold labelling of visfatin particles or small clusters consisting of number gold particles were demonstrated in the subcellular compartments of human colorectal HCT-116 mono‐ nucleated cells which were cultured in log phase of growth (Cy cytoplasm, N nucleus, Scale bars: a) 1 μm, b) 500 nm). Cells were chemically fixed, dehydrated and embedded in LR White resin [207]. Visfatin is an enzyme which overex‐ pression is correlated with poor prognosis in cancer patients. In this study, we tried to explore the association between visfatin distribution in subcellular compartments and increased apoptosis in cells treated with cytochalasin B. For fur‐ ther study, our aim is to optimise of HPF-FS protocols for cytoskeleton ultrastructure. Figure and legend adapted and

and legend adapted and changed from [207] under terms of the CC BY 3.0 license

**Figure 2a, b** Ultrastructural demonstration of immunogold labelling of visfatin particles or small clusters consisting of number gold particles were demonstrated in the subcellular compartments of human colorectal HCT-116 mononucleated cells which were cultured in log phase of growth (Cy cytoplasm, N nucleus, Scale bars: a) 1 μm, b) 500 nm). Cells were chemically fixed, dehydrated and embedded in LR White resin [207]. Visfatin is an enzyme which overexpression is correlated with poor prognosis in cancer patients. In this study, we tried to explore the association between visfatin distribution in subcellular compartments and increased apoptosis in cells treated with cytochalasin B. For further study, our aim is to optimise of HPF-FS protocols for cytoskeleton ultrastructure. Figure

**Figure 3a, b** Human colorectal HCT-116 were vitrified by HPF followed by FS and embedding in LR White resin. During FS process, cells were fixed with only 0,25 % GA. In comparison with cells prepared by conventional method (Figure 2), vitrified and freeze-substituted material show outstanding ultrastructure preservation. Even without OsO4 fixation, membranes are clearly visible, mainly rough endoplasmic reticulum and nuclear membranes. Nevertheless, further protocol

**Figure 3.** Human colorectal HCT-116 were vitrified by HPF followed by FS and embedding in LR White resin. During FS process, cells were fixed with only 0,25 % GA. In comparison with cells prepared by conventional method (Figure 2), vitrified and freeze-substituted material show outstanding ultrastructure preservation. Even without OsO4 fixation, membranes are clearly visible, mainly rough endoplasmic reticulum and nuclear membranes. Nevertheless, further

> a site of hydrogen accumulation [208] using quick-freezing preparation of TEM. Among the advantages of plastic sections, one is the section thickness; thus, comparatively large cellular volume can be analysed that was exemplified by microtubule cytoskeleton architecture in yeast [209], organellar relationship in the Golgi region of the pancreatic cells [210] or architecture of the caveolar system [211]. Besides, the reader can find many articles where either different techniques were

compared [55,91,96,212–216] or examples were collected in reviews [194,217].

Immunoelectron microscopy bridges the information gap between molecular biology and ultrastructural studies providing information regarding the function of the internal structures

level.

mesosome as

216 The Transmission Electron Microscope – Theory and Applications

changed from [207] under terms of the CC BY 3.0 license.

protocol optimisation is required.

optimisation is required

**8. Unity is strength: The hybrid techniques**

concentration of ions and the hydration status.

Cryosectioning according to Tokuyasu is one of the most reliable and efficient immunolocalisa‐ tion techniques for different types of sample. An inherent limitation of Tokuyasu cryosection‐ ing is mild chemical fixation at the beginning of the sample preparation. Thus, small molecules, including molecules of interests, may be dislocated or extracted during chemical fixation. Besides, it should be mentioned that process of chemical fixation is selective and results in pHrelated and osmotic changes in the different organelles [58]. Another restriction is intractabili‐ ty of samples, which contain a hydrophobic cuticle or a rigid wall, such as *C. elegans*, *Drosophila* or plants. Cryoimmobilisation should be used to overcome difficulties arising from chemical fixation during sample fixation. In spite of the all above-mentioned issues, one more should be considered – the nature of the antigen. Some antigens are sensitive on chemical fixation at room temperature or resin components and solvents and thus cannot be immunolabelled either in thawed cryosections or after cryoimmobilisation, freeze substitution and resin embedding. Therefore, the main aim was to introduce hybrid methods that combine the high-efficiency Tokuyasu cryosectioning labelling technique with an initial cryoimmobilisation step. Differ‐ ent approaches have been introduced with different results. The first attempts were taken by the group of Slot and Geuze [53,56] by combination of a frozen-hydrated cryosections with subsequent material fixation during thawing and after transfer to a grid. This method turned out to be unsuitable to routine use because, besides technical requirements and lack of reproducibility, only a small area of obtained sections has got desired morphological quality [56].

Another strategy to improve the antigenicity and ultrastructure preservation is rehydration method (RHM) based on cryofixation, freeze substitution and rehydration process before entering Tokuyasu cryosectioning and immunolabelling [56]. Sample after vitrification is dehydrated at low temperature with a substitution medium containing UA, glutaraldehyde and/or OsO4 and water to improve membrane contrast. After freeze substitution step, a rehydration process is carried out on ice. Rehydration step is necessary to enter the Tokuyasu procedure, i.e. sample embedding in gelatine, infiltration with high concentrated sucrose and freezing in liquid nitrogen. During dehydration step, additional chemical fixation is performed because fixation during FS step turns out to be insufficient. This approach resulted in excellent preservation of HepG2 cells, primary chondrocytes, cartilage and exocrine pancreases and immunolabelling efficiency comparable to Tokuyasu method. In the case of tested samples, authors suggested using the standard Tokuyasu technique because it is much easier, faster and allows the preparation of larger samples. The real power of the RHM methods was showed on *Arabidopsis* tissues, anthers containing pollen grains, *D. melanogaster* embryos and young adult nematodes. For these organisms, the RHM method was slightly modified [58,59] on dehydration step which was started at subzero temperatures. Obtained results were similar to the Van Donselaar et al. study: 1) water addition to FS cocktail appeared to be necessary for improved visibility of the membrane's bilayer structure and 2) sample fixation with osmium tetroxide, UA and glutaraldehyde during FS and post-fixation with glutaraldehyde at 0 ºC did not influence immunolabelling. Moreover, fixation-sensitive antigens were not inactivated, despite of using high concentration of fixatives. An additional benefit is usefulness of the hybrid techniques for fluorescence microscopy and CLEM due to the optimised ultrastructure preservation. Green fluorescent protein signal could be observed even after OsO4 treatment [58]. High lateral and axial fluorescence resolution can be obtained using thin cryosections; thereby, blurred signals are eliminated. Fluorescence-tagged antibody is much more sensitive than gold markers; thus, if fluorescence signal is not detected, then immunogold labelling is not worth to perform [59].

Hybrid techniques provide an alternative for worthlessness of vitrified cryosections in immunogold labelling. Recently, a novel hybrid technique called VIS2FIX was presented. First, vitrified material is cut into vitreous sections (VIS) and adhered to the EM grid. Next, the sections on the grid are fixed and brought to room temperature by means of FS and immuno‐ labelling. Vitreous sections are fixed (FIX) either by VIS2FIXFS method using a high-speed FS procedure with subsequent rehydration procedure or by VIS2FIXH approach based on waterbased frozen fixation, hence 'H' for hydrated [219]. Different combinations of fixatives were tested, including osmium tetroxide, glutaraldehyde, formaldehyde, acrolein and UA. The variety of mentioned fixatives and short time needed for sample preparation give huge possibilities in protocol optimisation for different antibodies and samples. The unique feature of VIS2FIXH method is fixation of lipids droplets, offering an interesting application in the lipidomics field. Although vitreous section is open structure and thus fixatives can penetrate it in high extent, the lack of embedding medium does not cause the material extraction. The high accessibility for the fixatives results in outstanding preservation of vesicles, particularly in the Golgi area and organelles. Further superiority over other techniques was proved through immunolabelling of both resin and aldehyde-sensitive antigens. As another option for making the impossible possible is vitrification of Tokuyasu-style immunolabelled sections, in brief VOS (vitrification of sections) technique. This approach was first time presented nearly a quarter of century ago by Sabanay et al. [220], and further afresh used in reconstruction of adhesion structures in tissues by cryo-ET [221]. In contrary to the described hybrid approach, in VOS technique, the first step is based on Tokuyasu sample preparation. The common steps are mild chemical fixation and cryoprotection in sucrose followed by immunogold labelling. After that, sample is re-vitrified in liquid ethane instead of treatment with methylcellulose and air-drying steps, as in the Tokuyasu technique [220]. The interesting thing is that the refrozen cryosections are free from cutting artefacts related with CEMOVIS procedure. Therefore, VOS technique provides meaningful 3D information on 300-400 nm thick sections. This method was used for 3D reconstruction of desmosomal adhesions in stratified epithelium and membranedense plaques and flanking caveolae in smooth muscle tissue [221]. The VOS technique is a valuable tool in fluorescence-based CLEM study, but together with Cryo-ET, its possibility to gain localisation of the targeted object in three-dimensional context is unique [222]. The advantage of VOS technique over standard Tokuyasu cryosectioning is that re-frozen cryo‐ sections are visualised by phase contrast results in more native sample preservation and overall better resolution. Moreover, structures invisible in Tokuyasu cryosections, i.e. cytoskeleton and ribosomes, should be possible to reveal with VOS.

In some conditions, chemical pre-fixation is an essential step prior vitrification process. This is true when safety consideration must be fulfilled, especially when pathogen organisms are the aims of the study. An aldehyde fixation step was introduced before vitrification of *Bacillus anthracis* spores [223]. Chemical pre-fixation step (CAF) followed by CEMOVIS made it possible to describe two new structures present in the spore. In some cases, chemical fixation is the only alternative for sample preservation. Biopsy specimen is particular situation, considering the place of sample collection where immediate processing of samples by HPF is rarely possible, i.e. a hospital or external laboratory. Thus, even sample is prolonged and stored in fixatives, subsequent processing by HPF followed by FS and polymer embedding results in better ultrastructure preservation compared to the conventional methods [224]. Some tissues are challenging specimen for ultrastructural preservation due to varied morphologies across its entity. Brain tissue integrity is degraded due to anoxia, unless chemical pretreatment is applied after the excision. The utility of this hybrid technique was demonstrated for variety of nervous system tissues. Aldehyde fixation prior to cryoimmobilisation and tailored FS protocol provides ultrastructural preservation superior to that obtained by conventional preparation and close to that obtained by HPF-FS protocol for tested samples [95]. Another challenging structure to preserve by HPF are stereocilia on the apical part of epithelial cells inside the inner ear. Preservation of the inner ear tissue using HPF is a challenging task, because the overall preservation of the sample was generally very good except for stereocilia. The explanation for this observation is unusual freezing properties of stereocilia actin bundles [96]. Authors considered different factors that could account for these properties, such as structure of the stereociliary actin bundles, high pressure during vitrification process and the treatment of the sample after freezing. Based on their research, authors were not able to unambiguously explain this phenomenon due to further study is required. Most importantly, chemical prefixation before cryoimmobilisation step resulted in preservation of cellular structure close to that prepared by HPF alone, and stereocilia actin bundle was preserved in a consistent manner.

### **9. Conclusions and outlook**

adult nematodes. For these organisms, the RHM method was slightly modified [58,59] on dehydration step which was started at subzero temperatures. Obtained results were similar to the Van Donselaar et al. study: 1) water addition to FS cocktail appeared to be necessary for improved visibility of the membrane's bilayer structure and 2) sample fixation with osmium tetroxide, UA and glutaraldehyde during FS and post-fixation with glutaraldehyde at 0 ºC did not influence immunolabelling. Moreover, fixation-sensitive antigens were not inactivated, despite of using high concentration of fixatives. An additional benefit is usefulness of the hybrid techniques for fluorescence microscopy and CLEM due to the optimised ultrastructure preservation. Green fluorescent protein signal could be observed even after OsO4 treatment [58]. High lateral and axial fluorescence resolution can be obtained using thin cryosections; thereby, blurred signals are eliminated. Fluorescence-tagged antibody is much more sensitive than gold markers; thus, if fluorescence signal is not detected, then immunogold labelling is

Hybrid techniques provide an alternative for worthlessness of vitrified cryosections in immunogold labelling. Recently, a novel hybrid technique called VIS2FIX was presented. First, vitrified material is cut into vitreous sections (VIS) and adhered to the EM grid. Next, the sections on the grid are fixed and brought to room temperature by means of FS and immuno‐ labelling. Vitreous sections are fixed (FIX) either by VIS2FIXFS method using a high-speed FS procedure with subsequent rehydration procedure or by VIS2FIXH approach based on waterbased frozen fixation, hence 'H' for hydrated [219]. Different combinations of fixatives were tested, including osmium tetroxide, glutaraldehyde, formaldehyde, acrolein and UA. The variety of mentioned fixatives and short time needed for sample preparation give huge possibilities in protocol optimisation for different antibodies and samples. The unique feature of VIS2FIXH method is fixation of lipids droplets, offering an interesting application in the lipidomics field. Although vitreous section is open structure and thus fixatives can penetrate it in high extent, the lack of embedding medium does not cause the material extraction. The high accessibility for the fixatives results in outstanding preservation of vesicles, particularly in the Golgi area and organelles. Further superiority over other techniques was proved through immunolabelling of both resin and aldehyde-sensitive antigens. As another option for making the impossible possible is vitrification of Tokuyasu-style immunolabelled sections, in brief VOS (vitrification of sections) technique. This approach was first time presented nearly a quarter of century ago by Sabanay et al. [220], and further afresh used in reconstruction of adhesion structures in tissues by cryo-ET [221]. In contrary to the described hybrid approach, in VOS technique, the first step is based on Tokuyasu sample preparation. The common steps are mild chemical fixation and cryoprotection in sucrose followed by immunogold labelling. After that, sample is re-vitrified in liquid ethane instead of treatment with methylcellulose and air-drying steps, as in the Tokuyasu technique [220]. The interesting thing is that the refrozen cryosections are free from cutting artefacts related with CEMOVIS procedure. Therefore, VOS technique provides meaningful 3D information on 300-400 nm thick sections. This method was used for 3D reconstruction of desmosomal adhesions in stratified epithelium and membranedense plaques and flanking caveolae in smooth muscle tissue [221]. The VOS technique is a valuable tool in fluorescence-based CLEM study, but together with Cryo-ET, its possibility to gain localisation of the targeted object in three-dimensional context is unique [222]. The

not worth to perform [59].

218 The Transmission Electron Microscope – Theory and Applications

The resolving power of TEM made it possible to visualise different objects at various resolution level ranging from angstroms for macromolecular complexes to nanometre scale for subcel‐ lular complexes and cells to micrometres for tissue gross morphology. Thus, electron micro‐ graphs have contributed significant understating of cellular structure and functions in physiological state as well as disease process. The sample preparation methods, however, still pose the main issue in biological TEM. It is important to realise that there is no single prepa‐ ration technique that could be applied universally. Appropriate choice of the preparation technique is determined by different factors:


The first aim is structural analysis. Sample architecture should be changed as less as possible for high-resolution analysis; therefore, cryoimmobilisation-based techniques are the best choice. Sample in near to native state can be prepared as follows:


For morphological and cellular structural biology studies, sample is prepared as follows:


The second aim of TEM is the immunocytochemical localisation of biochemically defined antigens within the cell landscape. Pure frozen-hydrated samples are useless for this task; thus, vitrified material is resin embedded after FS. Also specimen chemically fixed can be embedded in methacrylic resins for immunolocalisation study. Cryosectioning according to Tokuyasu is an alternative for both rare and sensitive to ethanol or methacrylates antigens. Tokuyasu technique excels in membrane contrast; hence, it is best suited for locating antigens in corre‐ lation to cellular compartments. On the other hand, resin sections have a larger area to analysis and better contrast that can be optimised for specific task. As it was mentioned in previous sections, hybrid techniques can offer a solution to antigens and samples difficult to fix for various reasons.

Another problem where preparation techniques pose the main issue is element distribution within analysed organism. Indeed, it is critical point to avoid and prevent diffusion of water and ions between cell compartments and outside the cell. To study element distribution within the cells and tissue, the best choice is CEMOVIS where sample remains at native state [62,63], at least in theory. Distribution of elements in resin sections obtained after vitrification and freeze substitution process are closer to natural state [64,226] in comparison with convention‐ ally prepared sample. Nevertheless, a conventional TEM is also used for analysis of ions in biological study [227,228].

**1.** A model organism. Vitrification should be method of choice for formidable to fix organ‐ isms such as plants or *C. elegans* where a hydrophobic cuticle, thick wall or starch granules slow down chemical fixatives and thereby induce artefactual morphological changes. **2.** When the time matters. Significant advantage of cryoimmobilisation over chemical fixation is the possibility to catch a dynamic cellular process at a known point and localisation of a rare event and structures, e.g. endomembrane transport in tissue culture cells, syncytial mitoses in early *Drosophila* embryos, nuclear division in *C.elegans* [103,225]

The first aim is structural analysis. Sample architecture should be changed as less as possible for high-resolution analysis; therefore, cryoimmobilisation-based techniques are the best

**•** By plunge freezing, dedicated to proteins, macromolecular complexes and viruses suspen‐

**•** By high-pressure freezing or self-pressurised rapid freezing for CEMOVIS or cryo-FIB, dedicated to samples thicker than 10 μm and by plunge freezing dedicated to small organisms, for analysis in sub-nanometre to nanometres range of macromolecular com‐

For morphological and cellular structural biology studies, sample is prepared as follows:

not influence on data analysis then conventional TEM is a good choice.

**•** By conventional sample preparation, dedicated for different samples, especially human tissue biopsies and whole organs of different animals. In some instances, this technique is more preferred, especially when the final contrast after cryoimmobilisation followed by FS is insufficient for further analysis. Good examples of this are synaptic vesicles in nerve cells and thylakoid chloroplasts [190]. If the extractions and/or distortions of the cytoplasm do

**•** By HPF/SPRF followed by FS combination; thereby, structural artefacts are limited. Resin sections give opportunities to obtain large-scale imaging at nanometre resolution.

The second aim of TEM is the immunocytochemical localisation of biochemically defined antigens within the cell landscape. Pure frozen-hydrated samples are useless for this task; thus, vitrified material is resin embedded after FS. Also specimen chemically fixed can be embedded in methacrylic resins for immunolocalisation study. Cryosectioning according to Tokuyasu is an alternative for both rare and sensitive to ethanol or methacrylates antigens. Tokuyasu technique excels in membrane contrast; hence, it is best suited for locating antigens in corre‐ lation to cellular compartments. On the other hand, resin sections have a larger area to analysis and better contrast that can be optimised for specific task. As it was mentioned in previous sections, hybrid techniques can offer a solution to antigens and samples difficult to fix for

Another problem where preparation techniques pose the main issue is element distribution within analysed organism. Indeed, it is critical point to avoid and prevent diffusion of water

or viral dynamics in cell [203].

**3.** The aim of the study and the object size.

220 The Transmission Electron Microscope – Theory and Applications

plexes and organelles within the cells

various reasons.

choice. Sample in near to native state can be prepared as follows:

sion, for atomic or near-to-atomic resolution study

Last but not the least is technical level of difficulty and equipment requirements. Cryotechni‐ ques are generally more technically demanding, including skills (e.g. cryosectioning of vitreous samples) and devices (e.g. cryo-ultramicrotome, cryo-EM, vitrification machines) in comparison with conventional TEM.

On the whole, the wide variety of available preparation techniques for biological samples enables to answer specific questions dependent on the study aims, model organisms and required resolution.

Considering the technological advances at EM field during the last decade, it is nearly impossible to predict future developments. Currently, some interesting solutions in both imaging and sample preparation at different level are under development, and presented ideas may significantly improve the ability to investigate and understand the world around us. Continuous software development for images acquisition, 3D reconstruction and further image processing and interpretation during electron tomography pushes the resolution limit down [229]. Another factor for improving resolution is new direct detection system with better quantum detection efficiency and high speed that allows corrections of beam-induced movements [230]. Further reduction of image distortions can be achieved by applying a holey carbon grid modified with graphene sheets [231]. An alternative way to improve final image of vitrified specimen is introduction of Zernike phase plate. This modification permits to record higher contrasted images with better resolution [232]. On the other side of scale of resolution, correlative light and electron microscopy has its own place. This modern approach is strongly explored in cell biology at various levels of resolution, and further progress is driven by scientific needs. Most recently, an integrated system for live microscopy and vitrification (MAVIS) was presented [233]. MAVIS combines a light microscope with a plunger to vitrify the specimen, and in this time lapse, imaging in a few second time resolution could be performed without the need for transfer step. An elegant solution called integrated light and electron microscope (ILEM) joins light and electron microscopes within one set-up and in this way greatly simplifies sample handling and navigation between the two modalities, therefore increasing final success of image correlation of both plastic sections and vitrified material [48,234]. Fluorescence signal during sample observation inside the column of ILEM is detected at dry environment; thereby, fluorescence marker should be picked carefully. Additionally, to avoid quenching and loss of signal, the en bloc staining, for example, via FS process, is a prerequisite for resin sections post-labelled with fluorescent dyes [48]. The ultimate goal of every microscopy technique which contribute to biology is visualisation of working life in as close to the native state as possible. Incompatibility of a liquid sample with the vacuum needed for electron microscopy seems to be apparent since electron microscopy of specimens in liquid is possible [235]. Another dream of electron microscopists has come true, but even so, appro‐ priate sample preparation is still a basic need.

The last decades have seen exponential technological progress that improves electron micro‐ scopes. Nevertheless, even if modern TEM will reach the electrons resolution limits, the sample preparation step remains a critical issue, limiting final achievements. Arguably, the optimisa‐ tion of sample preparation is key issue for the integration different microscopy techniques and joining data acquired at different-length scale into one view.

### **Acknowledgements**

The experiments were performed in the Department of Histology and Embryology in Zabrze, Silesian Medical University in Katowice using equipment financed by the Silesian Biofarma with infrastructure supported by POIG.02.03.01-24-099/13 grant: GCONiI – Upper-Silesian Center for Scientific Computation.

### **Author details**

Łukasz Mielańczyk\* , Natalia Matysiak, Olesya Klymenko and Romuald Wojnicz

\*Address all correspondence to: mieluk@gmail.com

School of Medicine with the Division of Dentistry in Zabrze, Medical University of Silesia, Department of Histology and Embryology, Zabrze, Poland

### **References**


[7] Kuo J, editor. Methods in Molecular Biology, Volume 369: Electron Microscopy. 2nd ed. Totowa: Humana Press; 2007.

The last decades have seen exponential technological progress that improves electron micro‐ scopes. Nevertheless, even if modern TEM will reach the electrons resolution limits, the sample preparation step remains a critical issue, limiting final achievements. Arguably, the optimisa‐ tion of sample preparation is key issue for the integration different microscopy techniques and

The experiments were performed in the Department of Histology and Embryology in Zabrze, Silesian Medical University in Katowice using equipment financed by the Silesian Biofarma with infrastructure supported by POIG.02.03.01-24-099/13 grant: GCONiI – Upper-Silesian

, Natalia Matysiak, Olesya Klymenko and Romuald Wojnicz

School of Medicine with the Division of Dentistry in Zabrze, Medical University of Silesia,

[1] Knoll M, Ruska E. Das Elektronenmikroskop. Zeitschrift fur Phys. 1932;78:318–39.

[2] Porter KR. A study of tissue culture cells by electron microscopy. J Exp Med.

[3] Lidke DS, Lidke KA. Advances in high-resolution imaging – Techniques for three-di‐

[4] Cortese K, Diaspro A, Tacchetti C. Advanced correlative light/electron microscopy: Current methods and new developments using Tokuyasu cryosections. J Histochem

[6] Hajibagheri MAN, editor. Methods in Molecular Biology, Volume 117: Electron Mi‐

mensional imaging of cellular structures. J Cell Sci. 2012;125:2571–80.

[5] Knott G, Genoud C. Is EM dead? J Cell Sci. 2013;126:4545–52.

croscopy. 1st ed. Totowa: Humana Press; 1999.

joining data acquired at different-length scale into one view.

222 The Transmission Electron Microscope – Theory and Applications

**Acknowledgements**

**Author details**

Łukasz Mielańczyk\*

**References**

1945;81:233–46.

Cytochem. 2009;57:1103–12.

Center for Scientific Computation.

\*Address all correspondence to: mieluk@gmail.com

Department of Histology and Embryology, Zabrze, Poland


[38] Fertig ET, Gherghiceanu M, Popescu LM. Extracellular vesicles release by cardiac te‐ locytes: Electron microscopy and electron tomography. J Cell Mol Med. 2014;18:1938–43.

[24] Kellenberger E. The response of biological macromolecules and supramolecular structures to the physics of specimen cryopreparation. In: Steinbrecht RA, Zierold K, editors. Cryotechniques in Biological Electron Microscopy. Berlin/Heidelberg:

[25] Chaplin M. Do we underestimate the importance of water in cell biology? Nat Rev

[26] Ho M-W. Liquid crystalline water, quantum molecular machines & the living tate.

[27] Pollack G. The Fourth Phase of Water – Beyond Solid, Liquid, and Vapour. Seattle

[28] Wineya M, Meehla JB, O'Toole ET, Giddings TH. Conventional transmission electron

[29] Dubochet J, Sartori Blanc N. The cell in absence of aggregation artifacts. Micron.

[30] Lucić V, Förster F, Baumeister W. Structural studies by electron tomography: From

[31] Massover WH. New and unconventional approaches for advancing resolution in bio‐ logical transmission electron microscopy by improving macromolecular specimen

[32] Stirling JW, Curry A, Eyden B, editors. Diagnostic Electron Microscopy – A Practical Guide to Interpretation and Technique. Chichester: John Wiley & Sons, Ltd; 2012. [33] Wągrowska-Danilewicz M, Danilewicz M. Current position of electron microscopy

[34] Cheville NF, editor. Ultrastructural Pathology: The Comparative Cellular Basis of

[35] Kanzaki Y, Terasaki F, Okabe M, Otsuka K, Katashima T, Fujita S, et al. Giant mito‐ chondria in the myocardium of a patient with mitochondrial cardiomyopathy: Trans‐ mission and 3-dimensional scanning electron microscopy. Circulation. 2010;121:831–

[36] Ding W-X, Li M, Biazik JM, Morgan DG, Guo F, Ni H-M, et al. Electron microscopic analysis of a spherical mitochondrial structure. J Biol Chem. 2012;287:42373–8.

[37] Popescu LM, Faussone-Pellegrini MS. TELOCYTES – A case of serendipity: The winding way from Interstitial Cells of Cajal (ICC), via Interstitial Cajal-Like Cells

in the diagnosis of glomerular diseases. Polish J Pathol. 2007;58:87–92.

Springer; 1987. pp. 35–63.

224 The Transmission Electron Microscope – Theory and Applications

Mol Cell Biol. 2006;7:861–6.

2001;32:91–9.

2.

Contemp Mater. 2010;1:117–23.

WA, USA: Ebner and Sons Publishers; 2014.

microscopy. Mol Biol Cell. 2014;25:319–23.

cells to molecules. Annu Rev Biochem. 2005;74:833–65.

preparation and preservation. Micron. 2011;42:141–51.

Disease. 2nd ed. Ames: Wiley-Blackwell; 2009.

(ICLC) to TELOCYTES. J Cell Mol Med. 2010;14:729–40.


[52] Tokuyasu KT. Use of poly(vinylpyrrolidone) and poly(vinyl alcohol) for cryoultrami‐

[53] Liou W, Geuze HJ, Slot JW. Improving structural integrity of cryosections for immu‐

[54] Griffith JM, Posthuma G. A reliable and convenient method to store ultrathin thawed cryosections prior to immunolabeling. J Histochem Cytochem. 2002;50:57–62.

[55] Bleck C, Merz A, Gutierrez MG, Walther P, Dubochet J, Zuber B, et al. Comparison of different methods for thin section EM analysis of Mycobacterium smegmatis. J Mi‐

[56] Van Donselaar E, Posthuma G, Zeuschner D, Humbel BM, Slot JW. Immunogold la‐ beling of cryosections from high-pressure frozen cells. Traffic. 2007;8:471–85.

[57] Griffith J, Mari M, De Mazie A, Reggiori F. A cryosectioning procedure for the ultra‐ structural analysis and the immunogold labelling of yeast Saccharomyces cerevisiae.

[58] Ripper D, Schwarz H, Stierhof Y. Cryo-section immunolabelling of difficult to pre‐ serve specimens: Advantages of cryofixation, freeze-substitution and rehydration. Bi‐

[59] Stierhof YD, El Kasmi F. Strategies to improve the antigenicity, ultrastructure preser‐ vation and visibility of trafficking compartments in Arabidopsis tissue. Eur J Cell Bi‐

[60] Dubochet J. The physics of rapid cooling and its implications for cryoimmobilization of cells. In: Allen TD, editor. Methods in Cell Biology, Volume 79. San Diego: Aca‐

[61] McDonald KL, Auer M. High-pressure freezing, cellular tomography, and structural

[62] Zierold K, Michel J, Terryn C, Balossier G. The distribution of light elements in bio‐ logical cells measured by electron probe X-ray microanalysis of cryosections. Microsc

[63] Nolin F, Michel J, Wortham L, Tchelidze P, Balossier G, Banchet V, et al. Changes to cellular water and element content induced by nucleolar stress: Investigation by a

[64] Volland S, Lütz C, Michalke B, Lütz-Meindl U. Intracellular chromium localization and cell physiological response in the unicellular alga Micrasterias. Aquat Toxicol.

[65] Moura Ramos JJ, Diogo HP. Are crystallization and melting the reverse transforma‐

cryo-correlative nano-imaging approach. Cell Mol Life Sci. 2013;70:2383–94.

crotomy. Histochem J. 1989;21:163–71.

226 The Transmission Electron Microscope – Theory and Applications

crosc. 2010;237:23–38.

Traffic. 2008;9:1060–72.

ol Cell. 2008;100:109–23.

demic Press; 2007. pp. 7–21.

Microanal. 2005;11:138–45.

2012;109:59–69.

cell biology. Biotechniques. 2006;41:137–43.

tion of each other? J Chem Educ. 2006;83:1389.

ol. 2010;89:285–97.

nogold labeling. Histochem Cell Biol. 1996;106:41–58.


[95] Sosinsky GE, Crum J, Jones YZ, Lanman J, Smarr B, Terada M, et al. The combination of chemical fixation procedures with high pressure freezing and freeze substitution preserves highly labile tissue ultrastructure for electron tomography applications. J Struct Biol. 2008;161:359–71.

[81] Koning RI, Zovko S, Bárcena M, Oostergetel GT, Koerten HK, Galjart N, et al. Cryo electron tomography of vitrified fibroblasts: Microtubule plus ends in situ. J Struct

[82] Medalia O, Weber I, Frangakis AS, Nicastro D, Gerisch G, Baumeister W. Macromo‐ lecular architecture in eukaryotic cells visualized by cryoelectron tomography. Sci‐

[83] Henderson GP, Gan L, Jensen GJ. 3-D ultrastructure of O. tauri: Electron cryotomog‐

[84] Kudryashev M, Lepper S, Stanway R, Bohn S, Baumeister W, Cyrklaff M, et al. Posi‐ tioning of large organelles by a membrane-associated cytoskeleton in Plasmodium

[85] Studer D, Michel M, Wohlwend M, Hunziker EB, Buschmann MD. Vitrification of ar‐

[86] Moor H, Riehle U. Snap-freezing under high pressure: A new fixation technique for

[87] Kanno H, Speedy RJ, Angell CA. Supercooling of water to −92°C under pressure. Sci‐

[88] Sartori N, Richter K, Dubochet J. Vitrification depth can be increased more than 10-

[89] Moor H, Bellin G. Cell and tissue: The influence of high pressure freezing on mam‐

[90] Hunziker EB, Herrmann W, Schenk RK, Mueller M, Moor H. Cartilage ultrastructure after high-pressure freezing, freeze substitution, and low-temperature embedding.I. Chondrocyte ultrastructure – implications for the theories of mineralization and vas‐

[91] Vanhecke D, Herrmann G, Graber W, Hillmann-Marti T, Mühlfeld C, Studer D, et al. Lamellar body ultrastructure revisited: high-pressure freezing and cryo-electron mi‐

[92] Van Belleghem F, Cuypers A, Semane B, Smeets K, Vangronsveld J, D'Haen J, et al. Subcellular localization of cadmium in roots and leaves of Arabidopsis thaliana. New

[93] Semmler K, Wunderlich J, Richter W, Meyer HW. High-pressure freezing causes structural alterations in phospholipid model membranes. J Microsc. 1998;190:317–27.

[94] Leforestier A, Richter K, Livolant F, Dubochet J. Comparison of slam-freezing and high-pressure freezing effects on the DNA cholesteric liquid crystalline structure. J

croscopy of vitreous sections. Histochem Cell Biol. 2010;134:319–26.

ticular cartilage by high-pressure freezing. J Microsc. 1995;179:321–32.

freeze-etching. Proc 4th Eur Reg Conf Electron Microsc. 1968;2:33–4.

raphy of an entire eukaryotic cell. PLoS One. 2007;2:e749.

fold by high-pressure freezing. J Microsc. 1993;172:55–61.

malian nerve tissue. Cell Tissue Res. 1980;216:201–16.

cular invasion. J Cell Biol. 1984;98:267–76.

Phytol. 2007;173:495–508.

Microsc. 1996;184:4–13.

sporozoites. Cell Microbiol. 2010;12:362–71.

Biol. 2008;161:459–68.

228 The Transmission Electron Microscope – Theory and Applications

ence. 2002;298:1209–13.

ence. 1975;189:880–1.


for section attachment and implementation of an anti-contamination glove box. J Struct Biol. 2010;169:219–25.

[123] Bouchet-Marquis C, Hoenger A. Cryo-electron tomography on vitrified sections: A critical analysis of benefits and limitations for structural cell biology. Micron. 2011;42:152–62.

[108] Yakovlev S, Downing KH. Freezing in sealed capillaries for preparation of frozen hy‐

[109] Sarkar P, Bosneaga E, Yap EG, Das J, Tsai W-T, Cabal A, et al. Electron tomography of cryo-immobilized plant tissue: a novel approach to studying 3D macromolecular

[110] Dubochet J, Lepault J, Freeman R, Berriman JA, Homo J-C. Electron microscopy of

[111] McDowall AW, Chang JJ, Freeman R, Lepault J, Walter CA, Dubochet J. Electron mi‐ croscopy of frozen hydrated sections of vitreous ice and vitrified biological samples. J

[112] Hodson S, Marshall J. Ultracryotomy: A technique for cutting ultrathin sections of unfixed frozen biological tissues for electron microscopy. J Microsc. 1970;91:105–17.

[113] Hutchinson TE, Johnson DE, MacKenzie AP. Instrumentation for direct observation of frozen hydrated specimens in the electron microscope. Ultramicroscopy.

[114] Chang J-J, McDowall AW, Lepault J, Freeman R, Walter CA, Dubochet J. Freezing, sectioning and observation artefacts of frozen hydrated sections for electron micro‐

[115] Al-Amoudi A, Norlen L, Dubochet J. Cryo-electron microscopy of vitreous sections

[116] Al-Amoudi A, Díez DC, Betts MJ, Frangakis AS. The molecular architecture of cad‐

[117] Bouchet-Marquis C, Dubochet J, Al-Amoudi A, Zuber B, Eltsov M. Cemovis cryoelectron microscopy of vitreous sections. In: Cavalier A, Spehner D, Humbel BM, edi‐ tors. Handbook of Cryo-preparation Methods for Electron Microscopy. Boca Raton:

[118] Griffiths G, McDowall A, Back R, Dubochet J. On the preparation of cryosections for

[119] Al-Amoudi A, Studer D, Dubochet J. Cutting artefacts and cutting process in vitre‐ ous sections for cryo-electron microscopy. J Struct Biol. 2005;150:109–21.

[120] Al-Amoudi A, Dubochet J, Gnaegi H, Lüthi W, Studer D. An oscillating cryo-knife reduces cutting-induced deformation of vitreous ultrathin sections. J Microsc.

[121] Michel M, Gnagi H, Muller M. Diamonds are a thief ' s best friend. J Microsc.

[122] Pierson J, Fernández JJ, Bos E, Amini S, Gnaegi H, Vos M, et al. Improving the techni‐ que of vitreous cryo-sectioning for cryo-electron tomography: Electrostatic charging

of native biological cells and tissues. J Struct Biol. 2004;148:131–5.

herins in native epidermal desmosomes. Nature. 2007;450:832–7.

immunocytochemistry. J Ultrastruct Res. 1984;89:65–78.

architecture of mature plant cell walls in situ. PLoS One. 2014;9:e106928.

frozen water and aqueous solutions. J Microsc. 1982;128:219–37.

drated sections. J Microsc. 2011;244:235–47.

230 The Transmission Electron Microscope – Theory and Applications

Microsc. 1983;131:1–9.

scopy. J Microsc. 1983;132:109–23.

CRC Press; 2008. pp. 259–89.

2003;212:26–33.

1992;166:43–56.

1978;3:315–24.


[149] Wang Q, Mercogliano CP, Löwe J. A ferritin-based label for cellular electron cryoto‐ mography. Structure. 2011;19:147–54.

[136] Wang K, Strunk K, Zhao G, Gray JL, Zhang P. 3D structure determination of native mammalian cells using cryo-FIB and cryo-electron tomography. J Struct Biol.

[137] Strunk KM, Wang K, Ke D, Gray JL, Zhang P. Thinning of large mammalian cells for cryo-TEM characterization by cryo-FIB milling. J Microsc. 2012;247:220–7.

[138] Fukuda Y, Schrod N, Schaffer M, Feng LR, Baumeister W, Lucic V. Coordinate trans‐ formation based cryo-correlative methods for electron tomography and focused ion

[139] Rubino S, Akhtar S, Melin P, Searle A, Spellward P, Leifer K. A site-specific focusedion-beam lift-out method for cryo transmission electron microscopy. J Struct Biol.

[140] Hayles MF, de Winter DAM, Schneijdenberg CTWM, Meeldijk JD, Luecken U, Per‐ soon H, et al. The making of frozen-hydrated, vitreous lamellas from cells for cryo-

[141] De Winter DAM, Mesman RJ, Hayles MF, Schneijdenberg CTWM, Mathisen C, Post JA. In-situ integrity control of frozen-hydrated, vitreous lamellas prepared by the cryo-focused ion beam-scanning electron microscope. J Struct Biol. 2013;183:11–8.

[142] Hsieh C, Schmelzer T, Kishchenko G, Wagenknecht T, Marko M. Practical workflow for cryo focused-ion-beam milling of tissues and cells for cryo-TEM tomography. J

[143] Wagenknecht T, Hsieh C, Marko M. Skeletal muscle triad junction ultrastructure by focused-ion-beam milling of muscle and cryo-electron tomography. Eur J Transl My‐

[144] Hayles MF, Stokes DJ, Phifer D, Findlay KC. A technique for improved focused ion beam milling of cryo-prepared life science specimens. J Microsc. 2007;226:263–9.

[145] Jun S, Ke D, Debiec K, Zhao G, Meng X, Ambrose Z, et al. Direct visualization of HIV-1 with correlative live-cell microscopy and cryo-electron tomography. Structure.

[146] Schwartz CL, Sarbash VI, Ataullakhanov FI, Mcintosh JR, Nicastro D. Cryo-fluores‐ cence microscopy facilitates correlations between light and cryo-electron microscopy

[147] Delgado L, Martínez G, López-Iglesias C, Mercadé E. Cryo-electron tomography of plunge-frozen whole bacteria and vitreous sections to analyze the recently described

[148] Sartori Blanc N, Studer D, Ruhl K, Dubochet J. Electron beam-induced changes in vit‐

and reduces the rate of photobleaching. J Microsc. 2007;227:98–109.

bacterial cytoplasmic structure, the Stack. J Struct Biol. 2015;189:220–9.

reous sections of biological samples. J Microsc. 1998;192:194–201.

beam milling. Ultramicroscopy. 2014;143:15–23.

electron microscopy. J Struct Biol. 2010;172:180–90.

2012;180:318–26.

232 The Transmission Electron Microscope – Theory and Applications

2012;180:572–6.

Struct Biol. 2014;185:32–41.

ol. 2015;25:49–56.

2011;19:1573–81.


of biological samples for immunoelectron microscopy. Histochem Cell Biol. 2012;138:167–77.

[176] Jiménez N, Vocking K, van Donselaar EG, Humbel BM, Post JA, Verkleij AJ. Tannic acid-mediated osmium impregnation after freeze-substitution: A strategy to enhance membrane contrast for electron tomography. J Struct Biol. 2009;166:103–6.

[163] Tomova C, Humbel BM, Geerts WJC, Entzeroth R, Holthuis JCM, Verkleij AJ. Mem‐ brane contact sites between apicoplast and ER in Toxoplasma gondii revealed by

[164] Humbel B. Freeze-substitution. In: Cavalier A, Spehner D, Humbel BM, editors. Handbook of Cryo-preparation Methods for Electron Microscopy. Boca Raton: CRC

[165] Monaghan P, Perusinghe N, Müller M. High-pressure freezing for immunocyto‐

[166] Steinbrecht R, Müller M. Freeze-substitution and freeze-drying. In: Steinbrecht R, Zierold K, editors. Cryotechniques in Biological Electron Microscopy. Berlin/Heidel‐

[167] Möbius W. Cryopreparation of biological specimens for immunoelectron microsco‐

[168] McDonald KL. Rapid embedding methods into epoxy and LR White resins for mor‐ phological and immunological analysis of cryofixed biological specimens. Microsc

[169] Sobol M, Philimonenko VV, Hozak P. Comparison of methods of high-pressure freezing and automated freeze-substitution of suspension cells combined with LR

[170] White DL, Andrews SB, Faller JW, Barrnett RJ. The chemical nature of osmium tetr‐ oxide fixation and staining of membranes by x-ray photoelectron spectroscopy. Bio‐

[171] Bohrmann B, Kellenberger E. Cryosubstitution of frozen biological specimens in elec‐ tron microscopy: use and application as an alternative to chemical fixation. Micron.

[172] Hawes P, Netherton CL, Mueller M, Wileman T, Monaghan P. Rapid freeze-substitu‐ tion preserves membranes in high-pressure frozen tissue culture cells. J Microsc.

[173] Walther P, Ziegler A. Freeze substitution of high-pressure frozen samples: The visi‐ bility of biological membranes is improved when the substitution medium contains

[174] Buser C, Walther P. Freeze-substitution: the addition of water to polar solvents en‐ hances the retention of structure and acts at temperatures around −60°C. J Microsc.

[175] Sobol MA, Philimonenko VV, Philimonenko AA, Hozák P. Quantitative evaluation of freeze-substitution effects on preservation of nuclear antigens during preparation

electron tomography. Traffic. 2009;10:1471–80.

berg/New York: Springer Verlag; 1987. pp. 149–72.

White embedding. Histochem Cell Biol. 2010;134:631–41.

chemistry. J Microsc. 1998;192:248–58.

py. Ann Anat. 2009;191:231–47.

chim Biophys Acta. 1976;436:577–92.

water. J Microsc. 2002;208:3–10.

Microanal. 2014;20:152–63.

2001;32:11–9.

2007;226:182–9.

2008;230:268–77.

Press; 2008. pp. 319–41.

234 The Transmission Electron Microscope – Theory and Applications


[201] Norlén L, Oktem O, Skoglund U. Molecular cryo-electron tomography of vitreous tissue sections: Current challenges. J Microsc. 2009;235:293–307.

[188] Kirschning E, Rutter G, Hohenberg H. High-pressure freezing and freeze-substitu‐ tion of native rat brain: suitability for preservation and immunoelectron microscopic

[189] Peschke M, Moog D, Klingl A, Maier UG, Hempel F. Evidence for glycoprotein trans‐

[190] McDonald KL. Out with the old and in with the new: rapid specimen preparation procedures for electron microscopy of sectioned biological material. Protoplasma.

[191] Durand A, Papai G, Schultz P. Structure, assembly and dynamics of macromolecular complexes by single particle cryo-electron microscopy. J Nanobiotechnology. BioMed

[192] Bartesaghi A, Matthies D, Banerjee S, Merk A, Subramaniam S. Structure of -galacto‐ sidase at 3.2-A resolution obtained by cryo-electron microscopy. Proc Natl Acad Sci.

[193] Schur FKM, Hagen WJH, De Marco A, Briggs J a G. Determination of protein struc‐ ture at 8.5Å resolution using cryo-electron tomography and sub-tomogram averag‐

[194] Bárcena M, Koster AJ. Electron tomography in life science. Semin Cell Dev Biol. 2009

[195] Ben-Harush K, Maimon T, Patla I, Villa E, Medalia O. Visualizing cellular processes at the molecular level by cryo-electron tomography. J Cell Sci. 2010;123:7–12.

[196] Ortiz JO, Brandt F, Matias VRF, Sennels L, Rappsilber J, Scheres SHW, et al. Structure of hibernating ribosomes studied by cryoelectron tomography in vitro and in situ. J

[197] Hohmann-Marriott MF, Sousa AA, Azari AA, Glushakova S, Zhang G, Zimmerberg J, et al. Nanoscale 3D cellular imaging by axial scanning transmission electron to‐

[198] De Jonge N, Sougrat R, Northan BM, Pennycook SJ. Three-dimensional scanning transmission electron microscopy of biological specimens. Microsc Microanal.

[199] Sousa AA, Azari AA, Zhang G, Leapman RD. Dual-axis electron tomography of bio‐ logical specimens: extending the limits of specimen thickness with bright-field STEM

[200] Wu JS, Kim AM, Bleher R, Myers BD, Marvin RG, Inada H, et al. Imaging and ele‐ mental mapping of biological specimens with a dual-EDS dedicated scanning trans‐

mission electron microscope. Ultramicroscopy. 2013;128:24–31.

localization of myelin glycolipids. J Neurosci Res. 1998;53:465–74.

2014;251:429–48.

2014;111:11709–14.

Oct;20:920–30.

2010;16:54–63.

Cell Biol. 2010;190:613–21.

mography. Nat Methods. 2009;6:729–31.

imaging. J Struct Biol. 2011;174:107–14.

Central Ltd; 2013;11 Suppl 1:S4.

236 The Transmission Electron Microscope – Theory and Applications

ing. J Struct Biol. 2013;184:394–400.

port into complex plastids. Proc Natl Acad Sci USA. 2013;110:10860–5.


[230] Grigorieff N. Direct detection pays off for electron cryo-microscopy. Elife. 2013;2013:2–4.

[216] Chlanda P, Carbajal MA, Cyrklaff M, Griffiths G, Krijnse-Locker J. Membrane rup‐ ture generates single open membrane sheets during vaccinia virus assembly. Cell

[218] Sendrowski K, Sobaniec W, Sobaniec P, Sobaniec-Lotowska ME. Ultrastructural study of hippocampal cortex neurons in an experimental model of valproate ence‐

[219] Karreman MA, van Donselaar EG, Gerritsen HC, Verrips CT, Verkleij AJ. VIS2FIX: A high-speed fixation method for immuno-electron microscopy. Traffic. 2011;12:806–14.

[220] Sabanay I, Arad T, Weiner S, Geiger B. Study of vitrified, unstained frozen tissue sec‐ tions by cryoimmunoelectron microscopy. J Cell Sci. 1991 Sep;100:227–36.

[221] Bokstad M, Sabanay H, Dahan I, Geiger B, Medalia O. Reconstructing adhesion structures in tissues by cryo-electron tomography of vitrified frozen sections. J Struct

[222] Bos E, Hussaarts L, van Weering JRT, Ellisman MH, de Wit H, Koster AJ. Vitrifica‐ tion of Tokuyasu-style immuno-labelled sections for correlative cryo light microsco‐

[223] Couture-Tosi E, Ranck J-L, Haustant G, Pehau-Arnaudet G, Sachse M. CEMOVIS on a pathogen: Analysis of Bacillus anthracis spores. Biol Cell. 2010;102:609–19.

[224] Venter C, van der Merwe CF, Oberholzer HM, Bester MJ, Taute H. Feasibility of high pressure freezing with freeze substitution after long-term storage in chemical fixa‐

[225] Caplan J, Niethammer M, Taylor RM, Czymmek KJ. The power of correlative micro‐ scopy: multi-modal, multi-scale, multi-dimensional. Curr Opin Struct Biol. Elsevier

[226] Boonrungsiman S, Gentleman E, Carzaniga R, Evans ND, McComb DW, Porter AE, et al. The role of intracellular calcium phosphate in osteoblast-mediated bone apatite

[227] Cloyd KL, El-Hamamsy I, Boonrungsiman S, Hedegaard M, Gentleman E, Sarath‐ chandra P, et al. Characterization of porcine aortic valvular interstitial cell "calcified"

[228] Bertazzo S, Gentleman E, Cloyd KL, Chester AH, Yacoub MH, Stevens MM. Nanoanalytical electron microscopy reveals fundamental insights into human cardiovas‐

[229] Vanhecke D, Asano S, Kochovski Z, Fernandez-Busnadiego R, Schrod N, Baumeister W, et al. Cryo-electron tomography: methodology, developments and biological ap‐

py and cryo electron tomography. J Struct Biol. 2014;186:273–82.

[217] Gan L, Jensen GJ. Electron tomography of cells. Q Rev Biophys. 2012;45:27–56.

phalopathy. Folia Histochem Cytobiol. 2013;51:31–7.

Host Microbe. 2009;6:81–90.

238 The Transmission Electron Microscope – Theory and Applications

Biol. 2012;178:76–83.

Ltd; 2011;21:686–93.

nodules. PLoS One. 2012;7:1–9.

plications. J Microsc. 2011;242:221–7.

tives. Microsc Res Tech. 2013;76:942–6.

formation. Proc Natl Acad Sci USA. 2012;109:14170–5.

cular tissue calcification. Nat Mater. 2013;12:576–83.


**Chapter 10**
