Research article

Construction and electrical control of ultrahigh-density organic memory arrays at cryogenic temperature

  • Mingjun Zhong 1, ,
  • Jie Li 1, ,
  • Yajie Zhang , 1, * ,
  • Xin Li 1 ,
  • Zhen Xu 2 ,
  • Qian Shen 3 ,
  • Xue Zhang , 2, * ,
  • Yongfeng Wang , 1, *
Expand
  • 1 Center for Carbon-based Electronics and Key Laboratory for the Physics and Chemistry of Nanodevices, School of Electronics, Peking University, Beijing 100871, China
  • 2 Spin-X Institute, School of Microelectronics, South China Uni-versity of Technology, Guangzhou 511442, China
  • 3 Key Laboratory of Flexible Electronics (KLOFE) and Institute of Advanced Materials (IAM), Nanjing Tech University, Nanjing 211816, China
*E-mails: (Yajie Zhang),
(Xue Zhang),
(Yongfeng Wang)

† These authors have equal contributions to this work.

Received date: 2023-02-27

  Accepted date: 2023-08-13

  Online published: 2023-08-15

Abstract

Investigation into the structural and magnetic properties of organic molecules at cryogenic temperature is beneficial for reducing molecular vibration and stabilizing magnetization, and is of great importance for constructing novel spintronics devices of better performance and scaling the device size down to nanoscale. In order to explore the possibility of fabricating molecule-based memory chips of ultrahigh density, two-dimensional close-packed molecular arrays with carboxylic acid molecules were constructed in the current work and the magnetic properties in a low-temperature scanning tunneling microscope were also investigated. The results demonstrated that each nonmagnetic molecule can be controllably and independently switched into a stable spin-carrying state at 4 K by applying a voltage pulse with atomic resolution. Benefiting from the small size of a single molecule as the basic storage bit, the two-dimensional molecular arrays allowing controllable electrical manipulations on each molecule can behave as a platform of memory chip with an ultrahigh storage density of ∼320 terabytes (Tb) (or ∼2500 terabits) per square inch. This work highlights the potential and advantage of employing organic molecules in developing future cryogenic information storage techniques and devices at nanoscale.

Cite this article

Mingjun Zhong , Jie Li , Yajie Zhang , Xin Li , Zhen Xu , Qian Shen , Xue Zhang , Yongfeng Wang . Construction and electrical control of ultrahigh-density organic memory arrays at cryogenic temperature[J]. Chip, 2023 , 2(3) : 100062 -5 . DOI: 10.1016/j.chip.2023.100062

INTRODUCTION

Operating electronic devices in a cryogenic environment helps to obtain excellent performance like reliable stability, low noise and high efficiency1-8. Compared with the RAM operating at room temperature, recent studies have demonstrated that cryogenic random access memory (RAM) is endowed with prominent advantages in read-write capability, reducing noise and power consumption9-11. However, scaling a cryogenic memory storage unit down to the nanoscale and further reducing the power consumption still remain challenging at present12. One of the potential solutions is to construct cryogenic memory chips with novel materials13.
Organic molecules are attracting extensive attention due to their structural stability, robust and tunable magnetic properties14-17. Moreover, their ability of self-assemble to well-ordered two-dimensional structures through hydrogen bonds18, coordination bonds and van der Waals force19-22. enables the creation of ultrahigh-density magnetic arrays. The combination of organic materials and cryogenic technology further brings promising opportunities to construct novel cryogenic magnetic memory chips with ultrahigh density and low power consumption. Scanning tunneling microscope (STM) and spectroscopy (STS) performed at low temperatures are powerful tools to investigate the structural and magnetic properties of organic molecules on the surface, facilitating the exploration of employing organic molecules in cryogenic storage devices. The two-dimensional molecular arrays formed by all-trans retinoic acid (ReA) on Au(111) have been reported previously by Wang et al.23-26, in which each ReA molecule can be transformed into a spin-carrying radical by applying a large voltage with the tip. Although the ReA molecule can behave as a promising candidate to construct magnetic memory arrays, its long polyene chain limits its potential to achieve ultrahigh storage density.
In the current work, (2E,4E)-3-methyl-5-(2,6,6-trimethylcyclohex-1-enyl)penta-2,4-dienoic acid (denoted as DiA) was chosen to construct close-packed molecular arrays on Au(111) due to its relatively shorter polyene chain. The topographic and spectroscopic features of DiA were investigated vis STM at a temperature of 4 K. Through the dehydrogenation process induced by an electric voltage pulse, a DiA which was originally in its nonmagnetic state could be switched to a spin-carrying state. The spin states are confirmed by STS and density functional theory (DFT) calculations and demonstrate a representative example of converting sequential molecules in the arrays into magnetic memory units, which implies the potential applications of organic molecules in constructing cryogenic information storage chips.

RESULTS AND DISCUSION

DiA molecules were thermally evaporated onto atomically flat Au(111) substrate with coverage of one monolayer at room temperature and conducted STM measurements at 4 K. A DiA molecule is composed of methyl-decorated six-membered ring and polyene chain and a terminal carboxy group, as illustrated in Fig. 1a. It was observed large-scale close-packed molecular arrays formed by DiA which were stabilized through intermolecular van der Waals force, as shown in Fig. 1b. In addition, from the close-up STM image in Fig. 1c, it is found that the basic periodic unit of the molecular arrays was a hydrogen-bonded DiA dimer in which two DiA molecules interacted with each other via the terminal carboxy groups23 as indicated by Fig. 1a and the dashed contours in Fig. 1c. The head part of the molecule appears as a lighter protrusion because of the two steric methyl groups. Compared to other organic molecules like metalloporphyrin and metal phthalocyanines whose potential in spintronics has been frequently reported27-29. DiA has a relatively smaller molecular size and is free of metal centers, benefiting the construction of high-density information arrays with the ease of manipulation. The ability of self-assembly via hydrogen bonding interactions also promotes its potential applications in spintronics. Moreover, a single DiA molecule in the arrays occupies less than 0.25 nm2, providing a nice candidate to perform magnetic state transversion and construct ultrahigh-density magnetic molecular arrays.
Fig. 1. Chemical structure and topographic images of DiA molecules on Au(111). a, Chemical structure of a hydrogen-bonded DiA dimer. b-c, STM images of self-assembled close-packed molecular arrays formed by DiA. The white dashed lines in c indicate a hydrogen-bonded DiA dimer. Image parameters: b, setpoint current It = 20 pA, sample voltage Vb = 1 V, image size 10 × 10 nm; c, setpoint current It = 125 pA, sample voltage Vb = 2 mV, image size 2 × 1.2 nm.
To convert the nonmagnetic molecular state of DiA to a spin-carrying state, the tip right above the six-membered ring part was located and applied a large negative voltage pulse at about −2.1 V. This could induce dehydrogenation and generate radical according to previous studies23,26. Fig. 2a shows a DiA dimer and its neighboring molecules in a molecular array and the voltage pulse imposed to the upper left DiA molecule. After the electrical operation, it is found the target molecule became much brighter than the others (Fig. 2b). This change in molecular topography is attributed to an electric field-induced dehydrogenation since a drastic current change was obscrved at around −2.07 V when the voltage was ramped from −2.05 V to −2.20 V (Fig. 2c), indicating a change in the molecular state. Notably, the pulsed electric field imposed to the molecule depends on the actual geometry of the STM tip, so the switching voltage can be slightly different from tip to tip. Differential conductance (dI/dV) measurements on the target molecule was performed to verify its spin state before and after the electrical manipulation and the results are exhibited in Fig. 2d. While the dI/dV spectrum of an unoperated DiA is featureless (black line), that of a switched DiA presents a profound peak near zero bias (green line), reminiscent of the Kondo resonance which was reported in earlier work23,24.
Fig. 2. Electric voltage manipulation of DiA and spectroscopic measurements. a, b, STM images of a segment of the DiA self-assembling arrays before and after applying a voltage pulse to the upper left DiA molecule. The white dashed lines in a portray the contour of a hydrogen-bonded DiA dimer. The black cross in a indicates where the STM tip was positioned to perform the voltage manipulation. Image parameters: a, b, setpoint current It = 55 pA, sample voltage Vb = 1 V, image size 1.2 × 0.9 nm. c, Change in tunneling current when ramping the voltage applied on a DiA from −2.05 V to −2.20 V. A drastic change emerges at around −2.07 V which hints to the occurrence of dehydrogenation. d, dI/dV spectra of the unswitched (black line) and switched DiA (green line) were measured on the black and green cross in a and b, respectively. Spectroscopic parameters: c, sample bias Vb = −2.05 V, setpoint current It = 480 pA; d, sample bias Vb = 20 mV, setpoint current It = 154 pA.
DFT calculations were further performed to better understand the switching process of the DiA molecule. Fig. 3a shows the optimized adsorption configuration of an individual DiA molecule on a Au(111) substrate consisting of three atomic layers, showing adsorption energy of −106.0 meV. The hydrogen atoms that are supposed to allow the dehydrogenation process are marked by blue arrows. By applying a voltage pulse atop the six-membered ring of DiA, one of the two hydrogen atoms can fall off and a neutral radical is formed, as shown in Fig. 3b. The adsorption energy of a dehydrogenated DiA molecule is calculated as −206.6 meV. The spin-resolved charge density distributions of a DiA radical in vacuum (Fig. 3c) and on Au(111) surface (Fig. 3d) reveal that the spin is delocalized and extends along the molecular skeleton. In addition, the polarity of the spin density of two adjacent carbon atoms tends to be opposite. The calculated spin magnetic moment of a DiA radical in vacuum is 1 μB, implying the total spin of the whole molecule is s = 1/2. In comparison, the calculated spin magnetic moment of a DiA radical on Au(111) is 0.67 μB, resulting from charge transfer between the molecular radical and the metal substrate.
Fig. 3. DFT calculations of the formation of DiA radical and corresponding spin-resolved charge density distribution. a, Schematic diagram of optimized adsorption configuration for an unswitched DiA on Au(111). The hydrogen atoms that are supposed to fall off by applying an electric voltage pulse are marked by blue arrows. b, Schematic diagram of a DiA radical on Au(111) formed by dehydrogenation. The remaining hydrogen atom bonded to the radical carbon is marked by a red arrow. c-d, Spin-resolved charge density isosurfaces of a DiA radical in vacuum and on Au(111) substrate, respectively. The isosurface value is 0.002 e/bohr3 and the blue and yellow isosurfaces refer to the spin-up and spin-down density, respectively.
To demonstrate the prospective of utilizing DiA molecules to build magnetic storage device, voltage manipulations were consecutively performed on sequential molecules in an array, as shown in Fig. 4a and 4b. Each molecule in Fig. 4a is marked by a grey circle with two arrows pointing in opposite directions to represent the nonmagnetic molecular state. Then the molecules in the left column were switched into magnetic radical state one by one using voltage pulses as introduced above, leading to a unified change in the topography (Fig. 4b), consistent with what was observed in Fig. 2b. With the aid of dI/dV measurements, it was verified that those molecules were successfully switched to spin-carrying states. Each switched molecule is marked by a blue circle with a single up arrow, implying the unpaired electron of the molecule.
Fig. 4. Magnetic storage arrays are formed by DiA molecules and controlled by electrical voltage manipulations. a-b, STM images of a DiA array before and after electrical voltage manipulations. Only the molecules in the left column were applied voltage pulses and became brighter afterward. The white dashed contour lines in a depict a hydrogen-bonded DiA dimer. The grey circle with two arrows in opposite directions refers to the nonmagnetic state of a DiA molecule, while the blue circle with a single up arrow indicates the spin-carrying state of a DiA radical after switching. Image parameters: a-b, setpoint current It = 55 pA, sample voltage Vb = 1 V, image size 1.8 × 1.2 nm. c, State transition of a byte composed of eight DiA molecular bits. Each grid refers to a molecular bit and the logic numbers ‘0’ and ‘1’ are used to represent the different states of a molecular bit, respectively.
Considering each molecule in Fig. 4a as a storage bit, logic numbers ‘0’ and ‘1’ can be used to represent the nonmagnetic and magnetic state of the molecule, respectively. Then the eight molecules compose a byte of a memory chip. The bits are all initialized to ‘0’ states (left panel of Fig. 4c), referring to the ‘00000000’ state of the byte without any interference. By applying electrical voltage pulses to the bits in the left column one by one, each bit can be converted from ‘0’ to ‘1’ and the byte is written as ‘10101010’ (right panel of Fig. 4c). Notably, two representative cases are only exemplified in Fig. 4c while all the 256 states of the byte can be ideally realized by performing controllable electrical manipulations with atomic resolution. Moreover, up to 3 × 1014 bytes are contained in a DiA molecular array of one square inch. In other words, a memory chip constructed by DiA can reach an ultrahigh density of ∼320 terabytes (Tb) (or ∼2500 terabits) per square inch. Notably, the information written into the molecule can be reversibly erased via electrical methods as previously reported23,24.

CONCLUSIONS

In summary, we successfully prepared large-scale two-dimensional organic molecular arrays using self-assembly strategies and investigated their potential in constructing cryogenic magnetic memory devices of ultrahigh density. The dehydrogenation can be induced and the nonmagnetic molecular state be switched to a spin-carrying radical state by applying a voltage pulse to the molecule. dI/dV measurements and DFT calculations further verify the magnetic molecular state after switching. By considering each molecule as a storage bit occupying a physical space of only 0.25 nm2, a DiA array composed of eight molecules forms a byte and a memory chip based on DiA arrays can achieve an ultrahigh density of ∼320 Tb per square inch. This work sheds light on the potential of employing organic molecules to construct ultrahigh-density magnetic memory arrays and provides useful fundamental reference for future study of novel molecule-based spintronics and RAM devices in cryogenic environments.

METHODS

STM experiments

All the STM experiments were performed in a commercial STM (Unisoku USM-1500SA) with ultrahigh vacuum and operated at 4 K. Atomically flat Au(111) surface was obtained by cycles of annealing and argon ion sputtering. A home-made k-cell evaporator was used for molecular evaporation. At a temperature of ∼150 °C, DiA molecules were thermally deposited onto a clean Au(111) surface which was kept at room temperature. dI/dV spectra were measured based on the lock-in amplifier technique with a modulation frequency of 967 Hz and modulation voltage of 0.1 mV. The feedback loop was disabled for spectroscopic measurements and electrical pulse manipulations.

DFT calculations

Density functional theory, including the spin-polarization effect, was performed for this research, as it is implemented in the Vienna ab initio simulation package (VASP)30,31. The core electrons were handled by the pseudopotentials constructed according to the projector augmented wave (PAW) method with plane-wave basis set32. The Perdew-Burke-Ernzerhof (PBE) form of generalized gradient approximation (GGA) was used to describe the electron exchange and correlation energy33,34. The electronic wave functions were expanded on a plane wave basis with a kinetic energy cut-off of 450 eV. The convergence criterion for the energy was 10−6 eV and that for ionic steps was 0.02 eV/Å. The first Brillouin zone was sampled with an Γ-centered k-point mesh. Au(111) substrate was modeled by a three-layer slab and the bottom two layers of the slab were fixed during the structural optimization. A vacuum layer of 12 Å was set between the neighboring slabs to separate inter-slab interactions along the z-axis.

MISCELLANEA

Acknowledgments This work is supported by the Ministry of Science and Technology (2018YFA0306003) and the National Natural Science Foundation of China (22225202, 22132007, 21991132, 21972002, 22172002, 21972067). DFT calculations are carried out on TianHe-1A at the National Supercomputer Center in Tianjin and supported by the High-performance Computing Platform of Peking University and Beijing Super Cloud Computing Center (BSCC) (URL: http://www.blsc.cn/).
Declaration of Competing Interest The authors declare no competing interests.
1.
Pauka, S. J. et al. A cryogenic CMOS chip for generating control signals for multiple qubits. Nat. Electron. 4, 64-70 (2021). https://doi.org/10.1038/s41928-020-00528-y.

2.
Gualco, G. et al. Cryogenic single-chip electron spin resonance detector. J. Magn. Reson. 247, 96-103 (2014). https://doi.org/10.1016/j.jmr.2014.08.013.

3.
Ishida, K., Tanaka, M., Ono, T.& Inoue, K. Towards ultra-high-speed cryogenic single- flux-quantum computing. IEICE Trans. Electron. E101.C, 359-369 (2018). https://doi.org/10.1587/transele.E101.C.359.

4.
Mhiesan, H., Hossain, M. M., Rashid, A. U., Wei, Y. & Mantooth, A. Survey of cryogenic power electronics for hybrid electric aircraft applications. In 2020 IEEE Aerospace Conference, 1-7. (IEEE, 2020). https://doi.org/10.1109/AERO47225.2020.9172807.

5.
Alam, S., Hossain, M.S., Srinivasa, S. R. & Aziz, A. Cr yogenic memor y technologies. Preprint at https://doi.org/10.48550/arXiv.2111.09436(2021).

6.
Aiba, Y. et al. Bringing in cryogenics to storage: characteristics and performance improvement of 3D flash memory. In 2021 IEEE International Memory Workshop (IMW), 1-4. (IEEE, 2021). https://doi.org/10.1109/IMW51353.2021.9439594.

7.
Alam, S., Hossain, M. S. & Aziz, A. A cr yogenic memor y array based on supercon- ducting memristors. Appl. Phys. Lett. 119, 082602 (2021). https://doi.org/10.1063/5.0060716.

8.
Resch, S., Cilasun, H. & Karpuzcu, U. R. Cryogenic PIM: challenges & opportu- nities. IEEE Comput. Archit. Lett. 20, 74-77 (2021). https://doi.org/10.1109/LCA.2021.3077536.

9.
Tannu, S. S., Carmean, D. M. & Qureshi, M. K. Cr yogenic-DRAM based memor y system for scalable quantum computers: a feasibility study. In Proceedings of the International Symposium on Memory Systems, 189-195. (ACM, 2017). http://doi. org/10.1145/3132402.3132436.

10.
Garzón, E., Teman, A. & Lanuzza, M. Embedded memories for cryogenic applica- tions. Electronics 11, 61 (2022). https://doi.org/10.3390/electronics11010061.

11.
Alam, S., Hossain, M.S.& Aziz, A. A non-volatile cryogenic random-access memory based on the quantum anomalous Hall effect. Sci. Rep. 11, 7892 (2021). https://doi.org/10.1038/s41598-021-87056-7.

12.
Yang, H. et al. Two-dimensional materials prospects for non-volatile spintronic mem- ories. Nature 606, 663-673 (2022). https://doi.org/10.1038/s41586-022-04768-0.

13.
Lian, H. et al. Metal-containing organic compounds for memory and data stor- age applications. Chem. Soc. Rev. 51, 1926-1982 (2022). http://doi.org/10.1039/ D0CS00569J.

14.
Zhao, A. et al. Controlling the Kondo effect of an adsorbed magnetic ion through its chemical bonding. Science 309,1542-1544 (2005).http://doi.org/10.1126/science. 1113449.

15.
Hiraoka, R. et al. Single-molecule quantum dot as a Kondo simulator. Nat. Commun. 8, 16012 (2017). https://doi.org/10.1038/ncomms16012.

16.
Xu, Z., Liu, J., Hou, S. & Wang, Y. Manipulation of molecular spin state on surfaces studied by scanning tunneling microscopy. Nanomaterials 10, 2393 (2020). https://doi.org/10.3390/nano10122393.

17.
Li, R. et al. Tuning the spin-related transport properties of FePc on Au(111) through single-molecule chemistry. Chem. Commun. 54, 9135-9138 (2018). https://doi.org/10.1039/c8cc02994f.

18.
Kumagai, H., Kepert, C.J. &Kurmoo,M.Constructionofhydrogen-bondedandcoordination-bondednetworksofcobalt(II)withpyromellitate:synthesis,struc-tures,andmagneticproperties. Inorg. Chem. 42, 3410-3422 (2002). http://doi.org/ 10.1021/ic020065y.

19.
Dong, L., Gao, Z. A. & Lin, N. Self-assembly of metal-organic coordination structures on surfaces. Prog. Surf. Sci. 91, 101-135 (2016). http://doi.org/10.1016/j.progsurf. 2016.08.001.

20.
Li, R. et al. Chemical way to tune spin excitation of magnetic atoms in Sierpi ´nski tri- angles. J. Phys. Chem. C 125, 5581-5586 (2021). https://doi.org/10.1021/acs.jpcc.0c11113.

21.
Zhong, M. et al. Spin-related electronic pathway through single molecule on Au(111). Chinese Chem. Lett. 34, 107813 (2023). https://doi.org/10.1016/j.cclet.2022.107813.

22.
Yuan, C. et al. Controlling metal-organic structure by tuning molecular size, sup- ported substrate, and type of metal. J. Clust. Sci. 32, 327-330 (2021). http://doi. org/10.1007/s10876- 020- 01791- x.

23.
He, Y. et al. Observation of biradical spin coupling through hydrogen bonds. Phys. Rev. Lett. 128, 236401 (2022). http://doi.org/10.1103/ PhysRevLett.128.236401.

24.
Karan, S. et al. Spin manipulation by creation of single-molecule radical cations. Phys. Rev. Lett. 116, 027201 (2016). https://doi.org/10.1103/PhysRevLett.116.027201.

25.
Gruber, M., Weismann, A. & Berndt, R. The Kondo resonance line shape in scan- ning tunnelling spectroscopy: instrumental aspects. J. Phys. Condens. Matter. 30, 424001 (2018). http://doi.org/10.1088/1361-648X/aadfa3.

26.
Bocquet, M.-L., Lorente, N., Berndt, R. & Gruber, M. Spin in a closed-shell organic molecule on a metal substrate generated by a sigmatropic reaction. Angew. Chem. Int. Ed. 58, 821-824 (2019). http://doi.org/10.1002/anie.201812121.

27.
Tsukahara, N. et al. Adsorption-induced switching of magnetic anisotropy in a single iron(II) phthalocyanine molecule on an oxidized Cu(110) surface. Phys. Rev. Lett. 102, 167203 (2009). http://doi.org/10.1103/PhysRevLett.102.167203.

28.
Kim, H., Chang, Y. H., Lee, S.-H., Kim, Y.-H. & Kahng, S.-J. Switching and sensing spin states of Co-porphyrin in bimolecular reactions on Au(111) using scan- ning tunneling microscopy. ACS Nano 7, 9312-9317 (2013). http://doi.org/10.1021/ nn4039595.

29.
Gopakumar, T.G., Tang, H., Morillo, J.& Berndt, R. Transfer of Cl ligands between ad- sorbed iron tetraphenylporphyrin molecules. J. Am. Chem. Soc. 134, 11844-11847 (2012). https://doi.org/10.1021/ja302589z.

30.
Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558-561 (1993). https://doi.org/10.1103/PhysRevB.47.558.

31.
Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169-11186 (1996). https://doi.org/10.1103/PhysRevB.54.11169.

32.
Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953-17979 (1994). https://doi.org/10.1103/PhysRevB.50.17953.

33.
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865-3868 (1996). https://doi.org/10.1103/PhysRevLett.77.3865.

34.
Perdew, J. P. & Wang, Y. Accurate and simple analytic representation of the electron-gas correlation energy. Phys. Rev. B 45, 13244-13249 (1992). https://doi.org/10.1103/PhysRevB.45.13244.

Outlines

/